Feedback about Hiragana 1 ~ 4
Almost finished Hiragana 1 ~ 4, since this is beta, I thought I'd give some feedback.
I dislike that words we learn in the Hiragana lessons and the Hiragana are combined. These should be separate lessons, even if these lessons would be very short.
Secondly, the choice of words are poor. Duolingo assumes a that you can intuit the language to a degree, but the words do not fit neatly into a single category.
We have a combination of nouns and adjectives.
The words we learn should be nouns and conjugates if possible, or at least fit a theme, like colors or numbers.
If a leaner reads ichi, they may not know what ichi is, but if they know that duolingo uses only numbers, they can make an educated guess. Right now, it could be a body part, a food, a number, a color or a greeting.
It is just really frustrating to look at a word and not be able to guess anything at all.
There is no way a word can be intuited unless it derives from a group you are familiar with. I can guess a lot of words in European languages due to Latin roots. I can guess a like 1/5 of words of Polish since my native is Russian/Ukrainian and they have high crossover. But when it comes to Japanese, it is used so far away, and there is next to no loan words. I was able to guess some Japanese words before I knew the meaning, but only due to prior exposure specifically to them (thx to黒子のバスケ for teaching me all major colors, though).
As for mixing up hiragana and vocabulary, I prefer current system. Hiragana is a tool, and the sooner you start using it, the sooner you master it. You don’t need pliers to hammer a nail, そうですか? You don’t need all sounds in alphabet for just one word. When you have to apply knowledge rather than just force it into your brain, you will remember it for much longer. What you suggest is learning raw facts separated into categories. In my opinion, learning a language is about developing some cognition within it first. For me, being able to say “ピーターはと申します. ウクライナ人です” is more useful than knowing all the color names.
Not saying current system is perfect, and I hate trying to figure out exact translations to make complex sentences accepted. Should improve as beta goes on, but for now we have to suffer the damage.
There is no way a word can be intuited unless it derives from a group you are familiar with.
This is not completely correct. Even if a word isn't derived from another, you can guess it through context.
If you are starting to learn your colors, and you know 白 and 赤, then you may be able to guess what 黒 is, even if you are unfamiliar with that Kanji. Even a wrong guess would be closer than trying to figure out the meaning of 目 from your knowledge of 赤, which is what duolingo is asking us to do.
In German, the first words you learn are Mann, Frau, Junge, Mädchen, Kind, Brot, Wasser, he, she, it and is. Man, woman, boy, girl, child, bread and water. People words, bread and water.
In Japanese, we learn, 1,2,3,4,6,7, good morning, yes, red, white, blue.
Hiragana is a tool, and the sooner you start using it, the sooner you master it.
There is no reason we cannot have the vocabulary in the beginning like it is, but broken into a separate lesson. Someone who is struggling with Hiragana doesn't need nonsense vocab words thrown into it. Likewise, someone who is struggling with the vocab, but who has a firm grasp of hiragana has no way to drill their vocab without spending a lot of time doing work on their hiragana.
I think it should be taught in order, AEIOU first, then Ka Ki Ku Ke Ko, Sa Shi Su Se So and onwards. I also think that hirahana should be taught first, with basic japanese only words and sentences learnt before Katakana is taught, so that people can tell the difference as some are similar.
I actually find it refreshing that they don't follow the same order that other languages do. Learning things in a less systematic way allows them to gently introduce the consonants, and also to build some words quickly and intersperse words with the characters, something I personally like.
Either way though these are not things I feel strongly about. I would much rather see the DuoLingo team prioritize things like bugs in pronunciation of certain characters, or addressing the lack of full Kanji in the course.
For now my biggest issue is that, for the questions of the form "which sound does this kana make", Duolingo automatically READS IT OUT, rendering the exercise entirely pointless, and thus hindering your ability to actually learn it (since you never have to recall it for yourself on those questions).
A similar flaw is present on the "match the pairs" questions, where it's possible to pair the kana to the sound just by pressing the kana first, hearing how it's pronounced, and never having to work it out yourself.
Yes, I know that this is because the underlying implementation treats "kana - pronunciation" like "foreign word - translation", but the underlying assumptions behind that are only a hindrance for the Japanese course.
I've messaged Duolingo several times via email and on Twitter but have as yet not received a response.
I have a lot of problems with the DuoLingo Japanese course, but I actually don't have a problem with any of the things you mention. I like that they start teaching you words as you're learning the kana, and I like that they teach you random words rather than having them all fit into a pattern.
DuoLingo has hover-over hints to help you learn what the words are...you don't need to guess. Just mouse-over the unknown word and it shows you the translation. This is how DuoLingo works in all languages...there's no need to get frustrated with this aspect of the course!