Just some anecdotal evidence to add to this. I’m at about 560 kanji “learned” right now, and I’m just starting to be able to read things without diving for the dictionary every sentence. Based on where I am, you might be able to get by with only the first thousand kanji, but 250 is nowhere near enough.
1 Like
I see this pop up a lot but can’t find anything - where did you get this number?
Not doubting you per se, by the way, just wondering if you have any additional information, mostly as to what that 80-90% figure refers to and what it’s based on.
Just from personal experience, which is worth squat in a scientific sense of course, you can fill out a lot of parts with significantly less than 80% of a text, I find, and things other than the text itself can provide context as well, so even if you end up not understanding all of what you’ve read you can still glean some new information and take that to whatever you read elsewhere (or even the same sentence) - is that not also learning from context? Or is the kind of learning you’re referring to different (i.e. not so much logical reasoning based, but more “intuitively understanding” what something means)?
2 Likes
Two links that may be of interest to you…
1 Like
Interesting reads, but both of those only seem to repeat those numbers, not so much support them (though I can only access the abstract for the first - maybe the actual article goes into them?). The second article even mentions the author has no idea where the 80% figure comes from.
Interestingly enough in the second article, I can’t for the life of me figure out what the unknown words in the 98% comprehensible text are supposed to be, but I can read and understand the 95% comprehensible text without problems, the 80% comprehensible text not so much but I can actually tell what quite a few of the nonsense words are supposed to mean. That doesn’t seem to really support the idea that more comprehension always means more/better learning from context - I learned precisely nothing from the 98% text after all, I have no idea what borgle and shnool are supposed to be. And I learned more “new words” from the 80% comprehensible text (bingle, maybe loopity - at least in vague terms, flid, weafling, joople, muchy, gutring, hunwre, assengle - on the second encounter, also giving a good guess about befourn, vickarn, parantle, gratoon) than the 95% comprehensible text (fossit, dowargle, quapen - also in vague terms at least, have a decent idea about farfoofle). The author claims 80% is “oh-so-frustrating” and leaves you clueless about the meat of the discussion, but… I don’t feel that at all, I could follow that quite neatly for the most part 
Wtf. Have you been spying on me? 
4 Likes
I was actually referring to myself. But, if the shoe fits… 
3 Likes
If you google the title of the paper you can find the pdf version. I haven’t posted it here directly because it may violate copyright laws.
1 Like
afaik it’s not scientifically based. Like you dont specifically need to know xyz in order to start reading or something.
It’s just that the most common top 1000 words usually cover approximately 80% of written text. Most people who learn through immersion or focus on input hypothesis usually try to learn that amount first before diving deep. just to make the starting process less annoying where you are forced to literally search every new word in the vocabulary.
obviously, you can start learning through context even with 100 words. It’s not a rule.
2 Likes
Ah, check. Yeah, that makes sense and fits my own experience a lot better - and I feel 1000 words is a good guideline for a foundation that lets you read with just the right amount of feeling like you’re bashing your head into a brick wall 
2 Likes
english furigana xD
Lmao they got me. I was reading the text and didn’t even realize there were nonsense words there. I thought borgle was a word I just didn’t know and shnooling was like when we play with noddle schoodle— to give an example Lol.
They got me good. 
2 Likes
I’d argue that with zero grammar, you’re not going to get most out of the context sentences during the lessons here on wanikani, which oftern helps with retention IMHO.
If you are fine without the context sentences, remember that often the solution is in the middle: I’d suggest reaching level 16 (standing to wkstats means 100% of N5 kanjis, 97% of N4 and roughly 50% of N3). In this way you can go through Genki I and II with a breeze for the grammar (or whichever grammar lessons for N5 and N4 for that matter), and then proceed in parallel with N3 grammar on something like Tobira and Wanikani to expand vocabulary.
At the end of the day, you know yourself. Personally I went for full grammar straight on, got to beginning of N3 and felt slowed down by my lack of vocabulary / kanji knowledge. So here I am stepping back and getting on par with vocabulary / kanji before tackling Tobira again. So verify if your lack in one of the two aspects is slowing you down and adjust your plan as needed.
4 Likes
That’s weird, it fits me too. I may have to rethink my image of the average WKer.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.