What I hate most about Wanikani? The context sentences aren't i+1

You see, this is where textfugu/etoeto/whatever should come in. Tofugu is losing the opportunity to monopolize the 日本語 teaching market.

“The WaniKani sentences are +1… if you follow our grammar platform at the progression recommended for each WaniKani level.”
“Our grammar platform has no new words… if you are on the recommended WaniKani level.”
“The SRS works best if you’re doing all Tofugu learning platforms together.”
“Here, these are the books for your current level.” * provides amazon afilliate link *
And so on.

This kind of thing would allow them to make possibly the best materials out there. Imagine a textbook that has no new words so you can focus on the grammar, yet still has deep text? That would be extremely motivating.

This would make them hundreds of thousands.

1 Like

I’m only level 6. Would you really like a level 6 to write a script for something they’re not fully competent in yet? Do you also prefer high school biology students performing your surgeries?

It’s not only inefficient for the WK community and myself to write a script at my current level, it’s nonsensical. WK has the resources and network to do so.

10 people doing just over 2 items a day ( i+1 sentences for Vocab and Kanji) will finish it in less than a year.

3 months if those same people complete around 9,3 items a day.

I think you don’t seem to realize just how much work would go into this? How will they handle grammar? How will they handle vocab not on the service? Non-Kanji vocab? What will intra-level progression look like? How do you deal with people potentially doing next levels lessons before covering the previous level vocabulary? Furthermore, WK is arranged from graphically simple Kanji to more advanced, but the terms covered can be a lot more advanced, as you probably know how do you go about covering this?

Further more, if multiple people need to work on this, how will they coordinate the known vocabularies, what method of oversight, how many times would you potentially need to reword or rethink a sentence so that it only fits the previously known words?

If you have a huge corpus of text it’s quite easy to go i+1 matching, but reverse-engineering i+1 sentences especially around a service that isn’t built with that in mind in the original order is a MONUMENTAL task.

If it’s something that you find important, feel free to take a shot at it, we’re more than happy to help you / correct you if you need it, but I don’t see why you’d require the level 60’s (who are just normal people such as you) to spend their free time doing something insanely complex as this?

5 Likes

But have you considered that if 2.8 people just write 3 letters every 45 minutes it will only take them 78% of a year.

4 Likes

I’m totally not against the team untertaking a couple of changes, and I’d very much suggest many of them myself. But a complete overhaul of the context sentences for more than 6000 words is not remotely within the scope of what I’m expecting. And to presume unpaid level 60s spendings months on writing a script for you instead of focusing on their studies feels just a bit audacious to me.

Like I said before, there’s no need to stress yourself about not understanding the context sentences right now. If you put in work outside of WK, you’re going to get better at understanding them eventually, and when that time comes, perhaps you’re not too bothered by the missing difficulty curve anymore. They haven’t done anything for me the first 20 levels either, though that wasn’t a dealbreaker for me because I knew the reason for this was me not knowing the language well enough yet, and I was already working on changing that. Ever since they’ve been incredibly helpful to me personally.

1 Like

I believe in you guys. :turtle::heart:

It may be near impossible but that means it’s possible.

I understand the complexity but surely if we lock in the default lesson ordering it should be feasible. Introducing new Kanji and Vocab after it’s implemented will be an issue but there’s enough of the aforementioned content on WK as it is and enough to work with.

I’m off from work the next few days, so I’ll do my research on creating scripts.

If I can make a separate tab called “i+1” or “Incremented Sentences” and make it work for the first couple of levels, maybe the community can contribute somehow.

It’ll be a neat feature.

Actually, I don’t exactly like the idea of introducing the Kanji to sentences after they have been learned. Vocabularies might be taken consideration of to some extent, but I generally disagree too (for the sake of technical difficulties).

With tools, it matters much less whether Kanji are learned or not.

At least I expect sentences to be translatable, and read correctly, with an aid of tools, like Yomichan or https://ichi.moe. (Or at least on par with Wikipedia / Japanese websites, which of course sentence level parsing, like ichi.moe, is error prone.)

IIRC, it’s only a problem for first 10 levels. (Readable for later-on learners and natives, but difficult for newcomers, even with tools.)

Google Translate Romaji guide

^ Yeah, the Romaji here is wrong, and I don’t think there are tools that can help.


Truthfully, example sentences shouldn’t be just automatically generated. They should be proofread too. And that’s pretty much the problem with Anime Context Sentence script. (On searching accuracy, and sometimes no results, even if there are.)

In the end, example sentences generally can be found in Goo, Weblio, ALC, or perhaps simply, Google. For the sake of convenience and reliability, probably try [Userscript] WaniKani JJ External Definition from Weblio (JP) and Kanjipedia

1 Like

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.