Re:The most frequent 777 characters give 90% coverage of Kan...
by andai on Saturday August 17, 2019 @04:04AM
(#20722826)
In English there's a lot of words you only see in books, you wouldn't pick them up by listening. So the only way to learn to read literature is to read literature, and tolerate a certain % of confusion. I learned most of those words through repeated in-context exposure, and eventually one will get stuck in my head later in the day and I have to look it up.
The downside of this is, when someone asks while they are reading, "hey what does [word] mean?", I often can't tell the precise definition, because I never looked it up, but I can usually read the sentence and understand it with context.
Re:The most frequent 777 characters give 90% coverage of Kan...
by Panoramix on Saturday August 17, 2019 @10:36AM
(#20724225)
It's much worse than that because kanji are not even words. The same kanji can appear in multiple words with absolutely nothing in common neither in meaning nor pronunciation. So the use of learning kanji is very limited vs learning whole words for example.
Re:The most frequent 777 characters give 90% coverage of Kan...
by nexus2045 on Saturday August 17, 2019 @04:01PM
(#20725987)
Heck, English is my most fluent language (though Chinese is my mother tongue) but I always have trouble understanding comments here on HN. I can understand every word and don't need to open the dictionary, but then I can't follow the logic / the point being debated / the thought being conveyed.
Re:The most frequent 777 characters give 90% coverage of Kan...
by wingerlang on Saturday August 17, 2019 @04:30AM
(#20722896)
How do you check your coverage? Or even how many words you know?
Re:The most frequent 777 characters give 90% coverage of Kan...
by bspammer on Saturday August 17, 2019 @09:29AM
(#20723925)
Just googling around, I found this test: https://www.arealme.com/vocabulary-size-test/en/
No idea about its accuracy, but as a native English speaker it's telling me my vocabulary size is 30k words, which sounds roughly correct.
Re:The most frequent 777 characters give 90% coverage of Kan...
by thwave on Saturday August 17, 2019 @06:38PM
(#20726883)
I've tried it now and got 22k, which seems not so bad for a foreigner ("Top 6.53% Your vocabulary is at the level of professional white-collars in the US!"), but I feel like I cheated: most of the more fancy English words are just misspelled Latin, and having even a modest Latin vocabulary (I'd don't think I know more than 4k Latin words) makes their meanings pretty obvious.
Re:The most frequent 777 characters give 90% coverage of Kan...
by xxxpupugo on Saturday August 17, 2019 @04:52AM
(#20722944)
There are certain tests floating around, like they will try 50ish quiz to estimate your vocabulary size.
Those are not set-in-stone science, so it is just an estimate, but I take about 10+ over those years, and the the number from those estimates seems consistent.
Coverage, there is some paper to track, I just googled it.
Take me as English learner for example. I would say I was only able to understand everyday English without too much of a hassle, after I acquired like around 10k words, which as I just checked had a coverage about 98%+.
Noted, it is still NOT enough, actually far from enough. Right now I believe I master around 15k to 20k words, by various estimates, and navigating English on the internet is like a charm, very little context switch in between with my native language.
Still, reading literature is huge undertake for me. I would still need to pardon myself about every once a page that if I stumped upon certain unknown words/phrases and can't move on before fully understands it, my pleasure of reading would be ruined. Such comprise frustrates me still, to this day. On the other hand, I will never have a second thoughts reading most cryptic novel in my own language, understanding might still be a challenge, but unlikely due to my insufficient vocabulary.