Incomplete Audio - Makes Certain Words Impossible To Learn Until They Appear In A Sentence
I notice that the Hebrew course, like the recently-launched Vietnamese course, lacks full audio. This means that I can't hear an individual word on hover-over, and perhaps more importantly, on the "pick the word with the image" multiple choice that is used to introduce a lot of words early on, there is no audio at all.
Furthermore, Hebrew isn't supported in Google's TTS, so I can't do what I do in Vietnamese or other languages, which is to copy and paste stuff into Google translate when I want to listen to it. I also hopped over to Forvo, a tool I frequently use to check pronunciations, and I was disappointed to find there are a lot of very common hebrew words that aren't listed there either, so while that does help in certain cases, it doesn't fully address the issue.
Because Hebrew doesn't have enough information to infer pronunciation fully from the writing (because the vowels aren't presented) this effectively makes certain words impossible to learn until they appear later in longer sentences that are spoken.
I have three questions:
(1) is there a time frame for getting full audio working?
(2) is there another external tool besides Google translate and Forvo that I can use to address these gaps?
(3) because I think this is such a big hindrance to learning the language, could the team re-work the exercises to eliminate all the "click the picture" type exercises until full audio is implemented? I notice that some words are introduced in the normal format, and these have audio...this seems like a good workable solution, it allows me to hear the words when they're introduced.
A very good observation, I was just about to write something similar. Ironically, the languages which need 100% audio the most in order to make proper sense of the pronunciation (Hebrew, Vietnamese, Irish) don't have it. How hard can it be to record several thousand individual words in addition to the thousands of sentences already recorded, or at least to provide a transliteration to the Roman alphabet or IPA where audio isn't available? Also, the script switching toggle is missing in action - they used it for the very easy to learn Cyrillic alphabet, yet they decide to omit it for the much more complex Hebrew abjad where the vowel information would be truly helpful. WHY???
Introducing new Hebrew words without audio (or at least a Roman alphabet transliteration) is almost as ridiculous as introducing new Chinese characters without audio. From a beginner's point of view, learning Hebrew words is almost as difficult as learning Chinese characters.
I so far haven't found it a problem - and the transliteration for the Cyrillic languages is horrible and makes my eyes bleed, so I haven't used it - but I do find it weird that it's available for the relatively simple Cyrillic but not for Hebrew.
My best guess is that while it is somewhat automatic for Cyrillic, for Hebrew I think it might require them to actually type in transliterations for each individual word, which I think would take a prohibitively long time.
I think for me personally, it's a Good Thing (TM), because it's forcing me to read the words, not rely on hearing them, but I have previous familiarity with the alphabet, which helps a lot.
I imagine not recording thousands of individual words is considered too expensive. I'm pretty sure it's nothing to do with picking one language over another, it's simply that courses which use a TTS get audio for everything, and those which use a recording do not. In some cases, the quality of the recording is such that it's decidedly a worthwhile compromise (Esperanto = fantastic audio, and the Hebrew so far has been beautifully clear), in others not so much.
(I have to disagree that new words in Hebrew are anything like as hard as learning Chinese characters 8-o Hebrew was challenging when I started learning it, sure, and it was frustrating to feel like I was plodding through a word letter by letter and still having to double check if I'd got it right, but IMO Hebrew is an order of magnitude easier than Chinese characters. Or heck, even the Japanese syllabaries. I mean, I'm sure it's a YMMV situation, but a slightly confusing abjad versus I have to be a freaking artist to replicate that? No contest!)
The team also made something great that really helps with the prononciation and overall learning : the words learned are on memrise : http://www.memrise.com/course/1031737/hebrew-duolingo/
That way you can learn them individually. All have audio. Definitely helped me.
Is there any way on Memrise I can play the sounds for the individual words, without going through the actual course? I really just want an external thing where I can look up a specific word and then play the audio for it, I don't really find Memrise's way of teaching effective for me and I think it actually creates bad habits in language learning so I avoid using it.
It is such an oversight for them to have released the course without pronunciations like that. Honestly, I would have rather waited until the course was complete in a more usable form for it to be released. As it is, I can't remember even the simplest words because I have nothing to connect them to in pronunciation and am therefore literally just trying to remember combinations of shapes with no attached sound.
Thank you so much! I tried it out and it seemed to get words that I know correct! This is exactly the sort of thing I was looking for!
It looks like this TTS is commercially available; is there any reason DuoLingo can't use it, or is it not quite at the level of quality they want?
it's true that hebrew isn't supported in text to speech but if you type in a hebrew word with google translate set to translate hebrew to english it will most often have nikud, which show the correct pronunciation.
1) this is a beta, so whenever the full version comes out is when i imagine they will have full audio. 2) hebrew dictionaries almost always have full nikud. having one would help a lot.
I am just working my way through very slowly. When I see a word for the first time with no audio, I just make up vowels for it in my head or use schwa sounds until I get to hear it being pronounced. My son is learning Urdu from a textbook, which is also lacking vowels, so there is a lot of guessing going on there too. But you can still read the language and improve from there.
I know this thread is two weeks old but I wanted to say I am very happy the course was released in beta even though it is not perfect yet. I started from scratch( no alefbet either) but I am learning lots and can use memrise to solidify the words and get the pronunciation.
I am not talking about hovering over words. I am talking about the listening exercises where you have to listen with no words. For beginners not used to the language you should have the option to have different listening speeds slow medium fast. Listening to some the phrases is like go ing to advanced level before finishing beginners level.
The reason for this is that we use a real recorded voice.
Therefore, the company we work with has recorded only full sentences (not individual words) - that's why you don't have voice when hovering sentences or in image exercises.
Honestly, there is nothing we can do about this voice thing, cuz that's what we are provided with.
However, I can tell you that i'm now starting a project that should ease those things.