Personally, my attempts to learn Japanese have been start and stop. I always search for a better way of doing things. Anyway, in January I used Rosetta Stone with resolution that I will use the Full Kanji mode. I'd practice each Kanji as it came up (big mistake in hindsight). Now that I discovered Heisig, I stopped with Rosetta (and other learning avenues) to concentrate more on learning the Kanji fully as suggested.

So, what are some opinions on Rosetta Stone? I like it personally when I first tried it. That you have ONLY Japanese to learn from helps in full immersion. Problems I have with it: Lack of basic structure explanation that would speed up understanding early on. It's stuck in one tense for verbs. The pictures are tied together in sets of four, with 10 sets per chapter. The review mode, though great in that you get to go back after a little miss, only encourages short term memory retention.

One big benefit, if you're into AJATT and SRS, is the PDF files of the entire course which is in English and Japanese. So, you can get the Kanji text in the question block, and put the Kana and English equivalent in the answer block. With over 10 thousand sentences, that can't be all bad.

I think Rosetta would be better if there was a way to get each picture, the Kanji sentence, pronunciation to it broken apart and placed in the SRS set-up. Rosetta's winning grace is that you have the picture to get the idea of the native sentence across. I would have the question be the Kanji, then the answer be the kana with the picture to see if I got both the pronunciation and the gist of the sentence down. MUCH better than relying on the english version above.

Rosetta Stone's greatest strength, as you mentioned, is that it does not provide English translations which completely changes how you have to approach it if you actually intend to learn anything.

I was first introduced to Japanese via Rosetta Stone and I can say wholeheartedly that it has wonderful concepts but its execution leaves much to be desired. Fortunately, I was able to access it for free via a library program as the actual program costs far too much for its lack of polish.

I, too, discovered that Rosetta Stone, while wholly awesome for short-term retention, did not, in of itself, actually teach or establish long-term habits, and this is greatest downfall, as it does not encourage the user to progress (and, in fact, does not measure his progress at all, period) or to review.

At the time, I was ignorant, but now I am equipped with the weapon of the gods. With SRS, one doesn't have to worry forgetting anything and Rosetta Stone could be quite useful. Additionally, it not only provides thousands of sentences in of itself, but provides all of those in a PDF, most conveniently.

Unfortunately, I have never actually paired Rosetta Stone up with SRS, but I'm sure that the results would be most splendid. Pretty soon (~four weeks), I'll be able to access Rosetta Stone again, and I'll start making sentence flashcards out of it.

I tried Rosetts Stone for Chinese and thought it was pretty terrible. It may be good for lots of phrases etc but actual dialogue practice? Also, it didn't break down the parts of the words and describe the tones accurately, this is vital for Chinese. You have to begin Chinese with an extremely accurate consideration of the tones. The course is used by the US army. Are the US army famous for their bi-lingual skills to accompany their kind reconstruction of countries they smash to pieces? i'm not a Rosetta Stone fan, as you can see. Pimsleur is my guru when it comese to starting languages from scratch.

Yes, Pimsleur is another one to look for. Ok, I've looked deeper into Rosetta. It uses .pct and .snd files for images and pictures. I already bulk converted the sound files to .mp3, so now I need to find a free bulk convertor for the graphic files to convert to .jpg.

Reason being, as Anki 3.0 will allow image and sound support, it's perfect for the job. After that, the only thing left is to get the sentences either typed in (ugh) or somehow get it from .pdf to a spreadsheet. The .pdf will not allow a copy/paste for some reason, so maybe I can sweet talk Rosetta into sending a spreadsheet format. With all that and learning the remaining kanji, I can jump right into the AJATT method without the annoyance of gathering sentences.

So, I will have nearly 8000 sample phrases, photos and sound files to work with Anki. In addition, my friend wants to learn Arabic (he already knows Mandarin and English fluently), so I'm gonna try to set up something similar for that language (using mnemosyne). Depending on the results, it may be worth a letter to the DoD on what is the most effecient method of distant learning a language.

Nukemarine Wrote:Reason being, as Anki 3.0 will allow image and sound support, it's perfect for the job. After that, the only thing left is to get the sentences either typed in (ugh) or somehow get it from .pdf to a spreadsheet. The .pdf will not allow a copy/paste for some reason, so maybe I can sweet talk Rosetta into sending a spreadsheet format.

For the pdf, you could try with a pdf editor -like Foxit or other- or, if the text happens to be in image format, you could use a OCR (Optic Character Recognition, if I'm not wrong) to convert it into editable text.

As for Anki, how do I add sound and images to my flashcards?

Transtic Wrote:As for Anki, how do I add sound and images to my flashcards?

According to the main thread, this was disabled in 0.3.0, but you could just use standard html in 0.2.8 for this effect.

Nukemarine, I have access to the Rosetta Stone CD's and I could bulk convert the files if given the proper tools: if you ever find a way to OCR or extract the text from the pdf's, please post how to do it as I would very much like to add some 8,000 phrases to Anki.

try searching for 'a-pdf restrictions remover'

The 15-day trial has no restrictions and might allow you to copy from the pdf

HTML input is not disabled in Anki 0.3.0 - it is the same as in previous Anki versions. I just said that I planned to add explicit support in 0.3.x :-)

Mighty_Matt Wrote:try searching for 'a-pdf restrictions remover'

The 15-day trial has no restrictions and might allow you to copy from the pdf

Anyone with a mac can do this easy enough.

Opening a pdf in Preview then highlighting the text you want to copy, selecting Preview > Services > TextEdit > New Window Containing Selection is all one needs to do.

Mighty_Matt Wrote:try searching for 'a-pdf restrictions remover'

The 15-day trial has no restrictions and might allow you to copy from the pdf

Thanks, it worked. Sadly, the copied text won't goto Word on my laptop so I'll try it at work later. I tried using mnemosyne which worked good. Here's what I've done

-Converted sound files using Cool Edit Pro
-Converted image files using Image Converter Plus
-Tested a couple of cards on Mnemosyne using its audio and image support. Initial card has only the Kanji sentence. The answer would be the photo, audio and kana (in hidden or inviso text). The reverse card has photo as the question, with the Kanji, audio, and inviso text kana. I manually need to create reverse cards as I don't want audio on the question card.
-After the format works, I just create all the files needed on a spreadsheet. For the Kana and Kanji, I have to go the long route and type them in manually. Perhaps this will be a good thing. However, if I can somehow get them from their pdf to a usable spreadsheet then I'll definately do it that way. 8000 slides is still a lot to do. For the spread sheet, I have to do alot of columns to recreate the needed formats so that mnemosyne reads everything the way I want it too.

The best layout would be the audio and kana be in their own "hint" blocks, with only the photos and kanji being the main cards, but you work with what you're given. On top of all that, I still have 1400 Kanji to memorize before I even begin attacking the Rosetta/Mnemosyne amalgams.

-EDIT- Well, there are .txt files that lists the romaji version of the pictures/audio on Rosetta's disk, so here's what I did. Opened in Word, did a find/replace and changed the @ to a line break (the @ were the seperators in the list). Copied that and went to a romaji to hiragana convertor site. Used the hiragana list on jwp to convert to kanji (that's the tedious part). Lastly I just paste the kanji into the excel spreadsheet. So, I now can start whenever. For now it's mnemosyne until Resolve puts audio/image support into Anki.

Now, since I can't post the photos nor audio nor the kanji sentences (well, legally), I can only post the method for others to sentence mine their own version of Rosetta. Wouldn't it be sad that I go through all this trouble and it turns out it's a lousy way to learn? Eh.

Nukemarine, you could always post it on box.net or something for all of us to test

I put the first unit onto the Anki site. First, create a Anki card list in Japanese format called "Rosetta", in there add two extra fields: Rosetta Number and Assistance. Sync that deck with account: Rosetta and password: Rosetta. Once that's done, you'll find that I suspended all but Unit 01, Lesson 01 (40 facts). Just sort the list by Rosetta Number and remove the suspended tag for a lesson as you go through each lesson.

If you have access to the sound and image files, you'll find it's easy to import them into Anki thanks to the Rosetta number. All image and sound have that number in their file name for very easy sorting. As theres 5 different image files for Rosetta, you may find the English images easier than the Asian images (when you see them, you'll understand), but the ideas still get across.

I have just finished getting the Kanji set-up for the first unit (400 sentences). I have them in Anki using .png and .mp3 format and it's working quite nicely. In addition, I added the English translation and kana readings (some typos in the kana), but it would work just as easy putting in Spanish or French or whatever. So on my phone, instead of a photo it'll show the English and Kanji sentence for cards. Thanks to the 0.3.5 update, I can keep both phone and computer in sync even though they display differently.

Still, I need to finish RTK as even at 1k kanji, half of the kanji in the first unit I do not have learned yet. If you end up using this I highly recommend Tae Kim's site to get the basic sentence structure down.

Who knows, it may turn out to be a useless experiment. I know you won't get the subtleties and slang with this. Of course, that's what AJATT is for. I am setting it up for a co worker to learn (native language is Cantonese), so there's be some data on how it works from another perspective. Something tells me this can be a successful experiment, but time will tell.

Here's a spreadsheet for the first unit of Rosetta

http://spreadsheets.google.com/ccc?key=p...nREQ&hl=en

It includes Kana, Kanji, English and Spanish, though any language can be added on to it making it fairly versatile.

For those that wanted to build a sample Anki (or other SRS) here's a link to the photos for Unit 0101 that you get for free anyway from Rosetta

http://picasaweb.google.com/Nukemarine/R...Lesson0101

Anything else, you'll have to strip your Rosetta program.

Again, you can also do Account: Rosetta Password: Rosetta at Anki and sync up with the Anki file there. Just put the photos in the Rosetta.Media and then put them in manually.

Due to the nature of RS there's going to be an incredible amount of redundancy in its content so you wouldn't want to use every sentence. Does the program ever get into dialogs or colloquial expressions? What about complex grammatical patterns?

I'm using Assimil Japanese with Ease and find that to be an excellent source for sentences. Here's a sample:

#48 そこから渋谷駅まで電車で行きます.
From there, we go by train until the Shibuya station.

#163 写真よりもっとうつくしい人ですよ。
She's more beautiful in person than in the photo.

#246 朝早いから、箱崎のエア・ターミナルまでリムジン・バスで行きます。
It's early in the morning, so I'll take the bus to the Hakozaki terminal.

#310 急いでいる時は汽車か飛行機で旅行した方が早いです。
When you're in a hurry, it's faster to travel by train or airplane.

#393 川崎先生によく似た猿が木の枝から枝へ飛び移っていました。
A monkey that looked a lot like my teacher, Mr. Kawasaki, was leaping from branch to branch.

Yeah, um, it does get into more complex areas as it goes on. However, you will not find natural conversation in this. It is in there briefly but I don't know how well it is. The english version seems passable enough, but that's never a guarantee. One the other hand, you get a large build up of sentence structure and vocabulary in context. So for now I do recommend using every sentence. While at first I was going to use this as a writing source (AJATT says to write your sentences), for this I'll treat it only as a reading and speaking source.

I know it seems stupid, but my training plan gets adapted fairly rapidly. It may turn out I abandon Rosetta altogether. I definitely don't care for the way it's presented for short term memory, but it combined with Anki may make it palatable. The annoying part has been finding the most efficient way to merge it and Anki. I AM NOT going to type in 6000 sentences that have mucho, mucho, repetitions.

Here's link to Unit 1 again.

http://spreadsheets.google.com/ccc?key=p...nREQ&hl=en

Here's Unit 02 (Japanese, English and Spanish) that I just finished editing.

http://spreadsheets.google.com/ccc?key=p...reVw&hl=en

With alot of RevTK under my belt, I'm noticing it's much easier to tell what the sentences actually mean. If you've done Rosetta you notice deciphering the image's context can be difficult. There's also some RTK3 Kanji spread throughout. As for "stilted dialogue", the sentences seem short enough to avoid that. Once again, time will tell.

Anybody currently putting these to use?

Nobody mentioned which version of RS they're talking about, but I have recently acquired Version 3, and from what I've read online (I don't have previous versions, so I can't say myself), the program has been much improved over earlier revs.

Background: My daughter, at age 10, got a wild hair to learn Japanese. We bought a few books, a dictionary, and RTK. Alas, neither of us kept up with it. Now, nearly five years later, she is planning to take it in school, so we're picking up again... and what a great resource koohii.com is, btw.

So anyway, here's what I've retained: nearly all the hiragana (but not instant recognition by any means), and about the first 20 or so kanji in RTK. We had never done any speaking or listening, so nothing to forget there :-) No vocabulary to speak of, and no grammar.

From this starting point, then, I am enjoying the Rosetta Stone v3 very much. Although I'm usually a learn-by-reading-everything engineer type, I was put off by my initial inquiries into grammar books. RS is gently introducing me to particles, and I'm actually pretty impressed by how they're managing to demonstrate grammar without any English. (Yes, it's probably very rudimentary grammar at this point, but it's farther than I'd gotten on my own.)

Each screen has the option to display kana, kanji, kanji with rubi, or (shhh, romaji). I'm displaying only kana until I've thoroughly assimilated all the phrases and sentences, and then going back and looking at them one time in kanji--not to add the kanji to my repertoire, but more just to expose myself to the mix of kanji and kana in a sentence (helping me recognize "word boundaries").

I'm trying to focus on speaking and listening in RS, as I don't have access to a native speaker, and I suspect very few JP podcasts are at my Dick-and-Jane level. As a bonus, my kana reading skills have noticeably improved in a very short time.

Without realizing that's what I was doing, I have been using some SRS-type principles in doing RS. I go through each little lesson then review it a day or so later, then go back to the ones I missed a few days after that... it's pretty plain to me that, especially for a rank beginner, doing each lesson only once or twice won't produce any gains. I've chatted with people online who complain that they get 100% on all the lessons and then 60% or 70% on the "milestone" (end-of-unit) test. Well, yeah!

There is only the same X number of pictures for each screen, and that's a limitation if you're repeating the lessons or parts of them many times. It would be nice if they had a pool of pictures for each one, so that if you reviewed you'd have at least a chance of seeing something different. However, they are definitely attractive, professional photos.

One trick I use (esp. when reviewing) is to listen with my eyes closed and imagine the picture *before* I open my eyes to look at the choices. If I don't do this, when I hear one word I recognize, then often I can pick the correct picture without fully understanding the sentence... once they say juusu, e.g., I can kid myself that I understood that there were several girl-children drinking it. Whereas when listening without seeing the photos, I pay more attention to the entire sentence.

Now, those of you who are further along in your studies are probably rolling your eyes, but I gotta say--in language learning (Spanish, French), I have always been fairly quickly proficient in reading but LOUSY at speaking and listening, but using RS, I feel confident that I will be able eventually to produce a few passable sentences, even if they're laughably simple.

As for kanji... I am committed to doing RTK, properly this time--did I mention we sorta skipped over flash cards and reviews last time? To my mind, these are two completely separate tasks, RS and RTK. I don't see much overlap, and I don't think one detracts from or hinders the other.

Except, of course, in the small matter of "hours in a day"...

Nobody mentioned which version of RS they were using because during the time when the last post was made in this thread, there was no V3.

It's nice that it's working out for you in some way. I purchased the entire V2 when I started, and while I was initially excited I soon became entirely unimpressed by the whole system and never got past around the fourth.. unit, or whatever it called them, the largest division. I still give them some credit though, as the vocabulary I learned there did stick.. at least for the nouns and adjectives. The verbs stuck too, but that was actually a BAD thing, since they were all in that damned ています form and I essentially had to un-learn them to actually learn them. But it helped me start getting used to Japanese sounds, and building confidence (even if it was artificial).

I've since seen the V3 French and such and found them much more impressive, so I'm considering looking into the V3 Japanese at some point...

Then again, I think I'm at a point now where it'd be largely a waste of time no matter how good it was. The fake immersion is great and everything, but when you can just use ACTUAL immersion...

QuackingShoe Wrote:Nobody mentioned which version of RS they were using because during the time when the last post was made in this thread, there was no V3.

Busted! I'm SO bad about not noticing that the last post in a thread is half a year old ...

I've taken a peak at the first disk for V3.

Pros: Has quick switching between Kana, Romaji, Kanji and Furigana (outstanding addition).

Better mixing of the training system. You go from Core Lesson, to testing Vocabulary, to testing Pronunciation, to testing Grammar, to testing Speaking/Listening. In later Lessons, it goes back to earlier lessons from time to time.

Better teaching format. Instead of just 4 screens, you know have from 2 to 8 photos shown. In addition, the way questions are asked change up. Sometimes it's pick the photo. Sometimes it's pick the sentence. Sometimes its fill in the blank. Sometimes it's audio reply. And when it introduces new words at times, it does it in a nice way: Show bread twice, then pick bread photo out of three choices. Show milk once, then pick milk out of three choices. Ask to pick water out of three choices (two of them are milk and bread).

Introduction of functional words right at the start. No Airplane or Elephant on the first lesson.

Actual changes in formality based on the situation. You don't see it in lesson 1 but it does come about.

Photos are MUCH, MUCH better this time around. It's easier to tell boys from girls in the photos.

It REMEMBERS how well you did on each lesson. This is good as V2, you were forced to get 100% to finish, but you had little idea how well you did in the lesson overall.

Cons: I think it's slow, but then I've been listening to Japanese ALOT. I remember the first time I did Rosetta Stone the pace seemed better, so maybe I'm just too advanced to give an unbiased opinion.

Still no SRS in the long term (may not be a problem). In V2, you automatically went back to missed segments after a time. Here, you're given an indicator and can go back on your own if you wish.

Anyway, disk 1 has 4 units of 4 lessons with about 2 hours of training per lesson (32 hours total?). I'd equate it to a lax semester of Japanese 101.

If you've done RTK, I'd say take some sentence from Rosetta (not all of them, just ones that have new material) and put them in Anki. Rosetta will not improve your writing, so this'll round it out.

Nukemarine Wrote:Personally, my attempts to learn Japanese have been start and stop. I always search for a better way of doing things. Anyway, in January I used Rosetta Stone with resolution that I will use the Full Kanji mode. I'd practice each Kanji as it came up (big mistake in hindsight). Now that I discovered Heisig, I stopped with Rosetta (and other learning avenues) to concentrate more on learning the Kanji fully as suggested.

So, what are some opinions on Rosetta Stone? I like it personally when I first tried it. That you have ONLY Japanese to learn from helps in full immersion. Problems I have with it: Lack of basic structure explanation that would speed up understanding early on. It's stuck in one tense for verbs. The pictures are tied together in sets of four, with 10 sets per chapter. The review mode, though great in that you get to go back after a little miss, only encourages short term memory retention.

um i dissagree with some of what you say.. i think that the reason you only see one verb tense is because you never got far enough into it.. i remember one of my lessons in it was specifically to teach different tenses of the verbs that were learned earlier.. and i also dissagree with the statement that it encourages short term memory not longterm.. but perhaps you had a different version than me.. i have the homeschool edition that gives tests and quizzes and tracks your scores so that your parents can see your progress at the end of the day to make sure you arent slacking off (it was me looking at it though tracking my own score).. the things you learn in the beginning are repeated in context first close together than only every so often.. however its deadly boring..

and i agree that it needs to have something to explain grammar better

you are brave though to try kanji mode only.. i certainly wasnt back when i used to use it all the time..

ugh ive got an older edition though.. i wish i had the money to upgrade.. i hear the new one has furigana.. i think it might be good for learning the kun yomi pronounciations after im done with rtk1/rtk2..

You are a master of necromancy.

People still give Rosetta Stone a second glance?

I miss my necromancer from D2 that got deleted when I couldn't play for a month .

Anyway... I tired RSv3 last year and it was great, I got pretty far... however. Nothing ever repeats so I would forget the previous section. I also tried v2 a while before that, but even though I almost finished v2 I've forgotten almost everything. I can still say something like "kodomo wa teeburu no shita ni imasu." which I'm sure will come in handy some day. Also I think I noticed that it doesn't use full kanji IIRC (some words are only in kana at some points, I did have it in full kanji mode all the time since I already knew Kana). Used alone I don't think RS is that great for Japanese, maybe Spanish, but it doesn't start to teach kanji or anything till like Unit 4. Thank the internet gods, I didn't pay for my version .

One good thing about RS was that I did a little bit of LifeMocha before I realized that their courses have some seriously bad Japanese (as told to me by native Japanese), a few native speakers that I befriended told me my pronunciation on the reading exercises was almost native. Thanks to shadowing the dialog in RS and using that voice recognition thing.

Nukemarine, you are unbelievably resourceful. This forum is lucky to have you around!

I am doing things in the AJATT order. I was toying with the idea of tackling Rosetta Stone after learning kanji-kana-grammar (currently at around 1600 kanji, with two weeks to go), to see how much more it would sink in.

I was told that I shouldn't bother with Rosetta, because there is so much free stuff that does the same thing. But it seems so neat and packaged and interactive. I am concerned with everything being in the same tense, though. You mentioned that. Is that right?

I was going to suggest keeping a desk calendar and do a manual SRS, writing down which Rosetta Stone chapters need to be studied which days at the appropriate intervals...but it seems that's not necessary with all the work you're doing to incorporate it into Anki!

Anyhow, I'll keep you posted if I jump to Rosetta Stone after finishing Tae Kim's grammar, and let you know how it goes. I am curious to see how truly effective it is in comparison to all the resources on iKnow.

am i the only one who's RS came with textbooks? i asked around on other forums and it seems not all of them come with the textbooks with each lesson written out for flashcard study..

ugh no the verbs arent all the same tense.. they are in the beginning but as you progress it changes.. however it doesnt explain them well enough.. you really do need a grammar suppliment.. i remember the lessons with verb tenses to be especially difficult for me back in the day when i was using the RS exclusively as my only learning tool..