In this look back at Smart.fm, I’ll cover the following topics:
- A review of the system
- How to attack it
- The system as a personal review tool
- The benefit of working through multiple goals simultaneously
A brief review of the system
By now, most Japanese learners have heard of Smart.fm (originally branded as iKnow). As a study tool, it’s not a bad system. Learners are presented with a small assortment of testing styles — Five-answer multiple choice (with one ‘none of the above’ option), ten-answer multiple choice, and fill-in-the-blanks based on meanings, readings, and Kanji. There’s also a dictation system, but it’s a separate application.
Pros:
- A flexible system open to user-created content
- Audio
- A professionally developed core system for beginner- and intermediate-level students
- Image-rich example sentences with appropriate audio
- A huge base of users constantly submitting content
- Personal progress tracking
- Scheduled study management across multiple “goals” (see below)
Cons:
- Some input problems on mixed Kana
- User-input can result in odd translations
- The system can lag at times
- Inputting sentences is a terrible process
- No options to control the testing format or how big a “bundle per round” is
How to attack it
The system is pretty straightforward. My only points of advice would be to pace yourself, focus on example sentences with the target words in context, and make it a habit to study from the system for at least 30 minutes a day.
By pacing yourself, you’ll remember more in a shorter amount of time, and you’ll need to cycle through the terms less frequently. I think this is one of the biggest problems that language students have. I understand the passion to drive through content in order to get to your fluency goal as soon as possible, but in reality that passion is actually slowing you down. Repetition is good to strengthen your acquired language foundation, but it’s not good when it’s out of necessity from having forgotten the words all together.
Seeing the words in context is another important part of building a strong L2 foundation. Seeing (and hearing) the words in an actual sentence helps you transition from the word as simply linked to your L1 lexicon, to the words being abstract blocks of memory, to gaining an intuition for usage.
Finally, it’s important to make the process habitual. As a general rule, if you can remember to study with the system for a week or two straight every day without fail, then you’ll most likely continue to do so out of habit. Also, instead of thinking of it as another daily task to complete, try and look forward to doing it everyday.
The system as a personal review tool
You can make use of the system’s flexibility to review content you’ve studied elsewhere. One of my particular favorite sources of study is the Harry Potter series in Japanese because of its progression in targeted reader age throughout the series. (Also, it’s very descriptive and has a lot of dialogue in it.)
Going through the Harry Potter series is a habit of mine, and when I come across terms I either don’t know, hardly know, or would like to use more, I put them into a specially made “goal” (like a study deck) that I’ve created, separating them into chapters, and then I review them after I’ve moved on to the next chapter. You can apply this idea to any content you’re studying, whether it’s a novel or a textbook.
The benefit of working through multiple goals simultaneously
I briefly mentioned this in the pros above, but being able to work through multiple “goals” and letting the system manage which ones need to be studied is probably one of the best parts of the Smart.fm system. The way I typically handle my studying is to enroll in several goals that have overlapping content in different orders. This is especially powerful when the associated sentences for target terms are different from the sentences in other goals you are also enrolled in.
Using this method, you get unexpected repetition, which is as close to “narrow-reading” as you can get without having to micromanage your own studies. Of course, on top of the Smart.fm content, I also use other “traditional” resources like textbooks (I don’t think textbooks are bad, especially when they’re written by experienced teachers), and slightly less traditional resources like novels, dramas, and sometimes even manga which also provide me with even more unexpected repetition.
What I’m getting at with this idea of multiple input of target terms is that this is what is considered ideal repetition. Many learners are under the mistaken impression that repetition means repeating the same things over and over again in an attempt to drill it into your mind. That’s a very antiquated idea of language study, and it’s extremely boring. Instead, what learners should be striving for is increased exposure for the purpose of unintentional repetition that will highlight words you’ve already encountered in a different context, and therefore build cognitive depth.
Take a break!
It’s just as important to take a mental break from the study process and let the information sink in. From time to time, pull back on pounding the knowledge into your head and sit back with some entertainment. Read a manga without a dictionary by your side. Don’t know a word? Skip it! You’ll figure it out later when you’re back on track with your studies. The most amazing part of this is that you’ll probably even remember encountering the word and not knowing it, which will make it even easier to remember upon later study.
Happy studying!
Related posts:
{ 1 trackback }
{ 25 comments… read them below or add one }
Some other things you didn’t cover:
The new beta tool for studying (which is pretty nice) doesn’t work on Linux -at all-. It seems to be fine on Windows and Mac OSX.
The old iKnow-style study tool is fine for Japanese, but when trying to input other languages, it doesn’t always work. Esperanto, for example, has special characters that aren’t on a US keyboard. OSX seems able to input the characters fine, but Linux fails, and the alternative way of typing Esperanto (with x’s after the modified characters) isn’t accepted. I did not test the add-on utility for Windows to type Esperanto, but I suspect it has similar issues to Linux because of how the input is done.
Admittedly, I used to switch between OS X and Windows, but these days I rarely boot up anything but Windows, so I wasn’t aware of any cross-platform issues from the beta tool. That’s definitely good information to keep in mind.
As a Linux user I just want to comment on the Linux thing above.
The smart.fm learning application is a JavaScript client and runs fine on most browsers, including Chrome and Firefox both of which run on Linux.
The input problems seem more related to your keyboard settings than anything else, but the application does have an inbuilt romaji-kana converter. For other languages/inputs you will need to have your own layout setup properly for accented characters or whatnot.
– disclaimer: I work for smart.fm
Thanks for dropping in and leaving a comment, Birkir. Since 한글 input works perfectly fine, I had assumed that it was probably an issue with keyboard input and not necessarily a problem with the system itself.
I’ve tried it, but I can’t say that I really know what I’m supposed to make of it. I tried a couple of the more popular goals and that just made me feel like I didn’t know really where I was at or where I was going… and I tried creating my own goal but that just made me feel that things would be easier and more customiseable if I just tried using Anki.
This is a good place to start off:
http://smart.fm/goals/19053
Those are the core 2000 words divided into 10 steps.
Anki’s downside is that there is no shared content (yet, at least last I checked). Damien (Anki’s creator) said he would implement some sort of system in the future, but it doesn’t exist yet. The Smart.fm goal I linked to above has professionally recorded voices.
One major problem I had with Anki is it starts to snowball as you build your deck and if you start missing out on study sessions or reducing your study session times.
Alex: “Anki’s downside is that there is no shared content…”
Would Anki’s shared deck functionality (where you can upload a deck to the Anki website and others can download it from inside Anki) not qualify as “shared content” or are you referring to something different here?
I haven’t used Anki for quite a while (because of the above mentioned snowball effect). The Shared Deck functionality isn’t quite the same because there’s no centralized database to search from (freely adaptable content). I think Damien had something planned in the future, but I haven’t been following its development lately.
I agree with your review mostly. I would add that in my experience the user generated content is all horrible. I’m only working with Japanese so I can’t speak for others. Initially Core 3000 step 3 was overwhelming for me. Too much hard vocab too fast. So I started signing up for other user-decks, various vocab themed or jlpt themed, et cetra. What I came across doing that is sentences with nonsensical translations or with translations that my Japanese wife says are completely wrong, questions that say one pronunciation but expect you to enter a completely different one to be correct, even questions in English that give away the answer.
While user-generated content is a great idea, smart.fm could review what is submitted a little better.
Actually, that brings up a really good point, David. I think Smart.fm should consider a quality ranking system – A very democratic system.
On another point – While it’s great to have such a convenient “tool” with a native-speaking wife (I don’t mean to call your wife a tool!), be a bit wary of native-speaker “intuition”. A lot of language is contextual, and I’ve heard from many native-speakers (in several languages, including English) that “X is not correct in Y-language.” On some of these occasions I’ve been able to show them that they’re wrong. One example in English is the word “so-so”. Some native speakers believe it’s “Japanese English”, but other native speakers use the phrase as well, so it may just be dialectical.
As another example, I asked my wife (she’s a native Korean speaker) if I she knew what “bird-watching” meant in English. She didn’t, so I told her it means “탐조” (探鳥, たんちょう in Japanese) – I had just learned the word in Korean, so I was really looking to impress her. She said that it isn’t a word people use in Korean. I doubted her, so we called up my father-in-law and she asked him if he knew what “탐조” was, and he replies, “Of course – It’s when you go out and look at birds.” If I had just gone off my wife’s response, I would have had the wrong answer, and she’s a smart woman, too!
The average Japanese speaker (or any native speaker, for that matter) doesn’t typically use a corpus to confirm their intuition, and sometimes they are looking at the same sentence from a different perspective. That’s the tricky part of studying isolated sentences.
Alex: “One example in English is the word “so-so”. Some native speakers believe it’s “Japanese English”, but other native speakers use the phrase as well, so it may just be dialectical.”
The phrase “so-so” (meaning something is not really good or bad, but rather just…ok) is definitely “normal” English (at least where I live in the SouthEast US is it). In fact, in Spanish class in high school, I recall being taught the Spanish equivalent of that particular expression (así así). So, if that is somehow “Japanese English” then it has apparently affected other Western languages as well.
That’s exactly my point. For me, it’s normal English. But I’ve heard the claim from an English teacher that it’s not “real English”, which I find dubious. I mean to say that the word may be dialectical in that perhaps it’s not used universally in the language. But to say that it’s not English is to take a very hard prescriptivist line in insisting that “my English is correct,” and not anyone else’s.
This isn’t only in English, although English is so diverse and widely spoken that it’s bound to have a ton of these sorts of examples. So my point is that asking for a native-speaker’s intuition and considering the response as the rule instead of a perspective is a mistake.
I completely agree. In fact, I’ve had to look up several British English words before as I hadn’t the slightest clue what they meant. If someone had asked me if one of those was the correct English word for that concept, I might very well have said “no” as I’d honestly never heard or read them previously.
My point was more that Spanish even seems to have the “so-so” phrase concept, so, if anything, I’d assume English got it from another Western language rather than from Japanese.
I don’t think the user generated content is all bad
You should be making your own goals, decks, whatever you want to call them and constantly revising them. The idea is rid yourself of the irrelevant, learnt or useless stuff.
Going from sentence level stuff to reading comprehension over longer passages is where your authentic text comes in. Alex the point you make about recognising the things you don’t know is a good one. You have to expose yourself to language as it is spoken and written, not endless loops of simple fragments.
I’d like Smart.fm to make it easier to create and edit items, let the users decide what is useful.
The thing is that I’m really about 漢検５級/JLPT2級 as far as my vocab goes. So, I tried the Intermediate Japanese goal, but that just seemed a mixed bag of compounds, kanji and sentences. Really I’d much rather learn my vocab from the kanji down rather than being given a lone random ‘intermediate’ compound with unknown kanji in it and told to learn it. Going with the kanken levels I can use different materials to help me along and have actual goals to aim towards, and I can supplement this with vocabulary that I come across naturally in my reading, but going with smart.fm I just feel like it’s trying to force feed me in some arbitrary manner.
Nice little review. As I mentioned back on the NihongoUp review at my blog, I agree with you about narrow reading and am intrigued by Smart.fm, haivng not used Anki for a while (it can get quite tough to keep up with). But as others have mentioned above, how do you go about finding quality decks, and more inportantly, how can you be sure the decks the user has created are correct when they are more than likely not native speakers themselves? Any particular decks you recommend other than the ones you mentioned?
Here’s a (short) list of quality decks:
Japanese Core 2000 Steps 1 – 10
Intermediate Japanese Steps 1 – 12
Hiragana Time’s Articles
Other than those, if you’re into learning from Japanese music videos there are a few available.
You can also use it to review RTK, but there’s a learning curve to the process. Mainly, I use it to create my own review lists of content from tangible books.
I’m sure we’re all hoping for Smart.fm to produce more high-quality Japanese content, but beggars can’t be choosers.
I should have mentioned that another great aspect of Smart.fm is the ability to open your goals for collaboration. That way they are constantly refined by the crowd.
Here are some I’m collaborating on. I wrote about opening goals for collaboration in a post in July of last year. Here are some more goals I’ve created.
I reckon some of your unusual katakana words would make a great goal Michael. How about it?
You may want to check out some of my decks… They are fairly basic but good for new learners.
Japanese Color Names
15 Essential Japanese Drinks
Human Japanese – Open deck with tons of vocabulary from Human Japanese
Michael,
Here’s the odd thing, many of the learners at RevTK forums loved the content of iKnows official decks (Core 2k, Core 6k specifically), but many of us got turned off by the reviewing system. The user created decks were junk usually, thanks in part because any deck created was by default a public deck. As it took effort to hide a deck, most were just kept public. Thus, the signal to noise was horrible.
As for the quality decks of Core 2k and 6k (Core 6k is now called Intermediate Japanese), it has the problem listed in the above comment in that they’re in either random or “alphabetical” order. What some users on RevTK forums did was create a program that sorted the sentences based on Kanji (more specifically, 2001.Kanji.Odyssey order). That means you don’t see sentences (and by proxy, vocabulary words) with random kanji appearance. Instead, kanji are gradually introduced in an intuitive order making for a pleasurably learning experience based off reports of those that used the new order. Of course, to use these sorted sentences, you have to use Anki unless Smart.FM releases a resorted deck.
Also, there was a RevTK group effort to put vocabulary from 2001.Kanji.Odyssey into large decks (which should count as quality). Do a search for “CosCom” or “KO2001″ on Smart.FM to bring up the 3500 vocabulary lists derived from books 1 and 2. There’s also a project to get vocabulary from book 3 in there.
I have tried smart.fm again with Thai. It’s really very useful. I typically don’t learn from lists generated by other people. But in this last trip to Thailand, I found a few words I’d learned popped into my head. Must be the review games — I got addicted. lol
I’ve started to base my assessment on compiled lists on Smart.fm on the amount of sentences provided in the goal. Before I start studying anything using the system, I check the quality of examples they provide.
Glad to hear you found the system useful, and that the content actually stuck in your memory!
When you said that you could work on multiple goals and the system would tell you which goal you should be studying, how do you get it to do that? My goals are always listed in order of when they were last studied, not based on how much of it I’ve mastered or any other factors that should be taken into consideration, in the way that the actual words reviewed while studying are chosen based on multiple factors like how well you’ve remembered it previously, etc. Please help, because that sounds very useful/helpful.
I didn’t do anything special – That’s just the way it was for me by default. Maybe you can monitor it a bit more. I don’t think there’s any special activation option.
Ok, maybe I just misunderstood what it was doing. Thanks a lot.