Reverse (English to Chinese) cards in Anki / SRS

[werewitt]

A question for you people who use Anki or another SRS. When I used Anki to drill vocab for dead languages, I obviously had a one-card model, foreign to English. In my Chinese deck, I have both reverse and forward cards (just added a char to pronunciation card as well). 

 

(1) Any opinions on the usefulness of English to Chinese cards? Given all the synonyms it's not too useful. Some sort of E-Ch production is obviously useful - I am wondering whether they are worth spending time on, given I could spend it on something else, like learning more sentences. Or something like this? http://forum.koohii.com/thread-10759-post-190214.html#pid190214

 

(2) If drilling sentences, is it better to learn those sentences Chinese to English or both ways?

 

My current goal is more than just reading, it includes some passable level of speaking as well. And of course Anki is not my main/only tool.

[Lumbering Ox]

I did RTK Keyword-Kanji. There are a lot of Keywords that are basically the same and it was a real bohica to keep them straight.

What I did was added a note to the key word. I can't remember an example as it was a while ago but suppose

A and B mean the same and have two different Kanji. Normally I'd remember both Kanji but not have a clue which was the right one.

 

Card A would be A -not [some part of the story for B]

Likewise Card B would be B - not [Some part of the story for A]

If I knew how to type Kanji at the time I'd probably put the similar Kanji in the keyword but I didn't.

 

Would such a system work for E to C vocab decks?

 

I have read that having to recall a second language word is more taxing to the fresh tasty brains than just trying to recognize second language.

Much like the difference between active [what is the word for???] vs passive [oh yeah I know that word].

Passive vocab seem to stick much easier. I never have to think about what a word means in English even though I am old. If there is a word I don't know, it is a word that is never used and try as I might, I just don't have a clue. OTOH I am always having to come up with a word when speaking, sometimes annoyingly common words to the point that at times I sound like Jean Cretian or a Mafia guy. "Them guys who do those things"

 

 

If at all possible, I'd prefer to use E to C decks unless someone tells me that it is just not doable with the hack above or any other hack.

[NinKenDo]

I would only do this with a tool like this. Which allows you to focus less on remembering an exact 1-to-1, fixed translation that is right or wrong, and focus on using syntactic units which you know to construct sentences appropriately.

 

There is no one way to translate one thing in English, not even something as simple as "hello" or "me", so trying to remember precise 1-to-1 translations is not only frustrating and inefficient, but counter productive. You should be learning to produce the same English meanings with a variety of Chinese syntactic units, and that can only be probably achieved when you have some way of letting yourself know which units you are to use, without simply giving yourself the answer in the process.

[werewitt]

11 hours ago, NinKenDo said:

 

I would only do this with a tool like this.

 

Weird, I looked at the plugin (Sentence Scalpel: Stand Alone Complex) this morning and noted as worth trying, yet now it has disappeared. 

It looks like that toolset is in flux, https://ankiweb.net/shared/info/1956177600

[Wurstmann]

17 hours ago, werewitt said:

Some sort of E-Ch production is obviously useful

 

I think it's only useful if you want to become a translator or interpreter. 

[werewitt]

11 minutes ago, Wurstmann said:

I think it's only useful if you want to become a translator or interpreter.

So you're saying it's does not help at all in attempting to speak and you never learned production, only recognition (in SRS terms)? Before you had an opportunity to for immersion.

[trevorld]

Quote

(1) Any opinions on the usefulness of English to Chinese cards? Given all the synonyms it's not too useful.

 

Don't simply click a button and create "reverse" cards for English -> Chinese for all your Chinese -> English cards.  However creating a new limited set of cards of Meaning -> Chinese might not be a bad idea (as suggested in Fluent Forever sometimes it is better to use pictures instead of an English word because ideally you want to go straight from meaning -> Chinese and not necessarily meaning -> English -> Chinese).  Keep in mind the following:

 

1.  It is more efficient to acquire a larger passive vocabulary then active vocabulary so the number of Meaning -> Chinese cards (i.e. active cards) should optimally be much smaller then the number of "Chinese -> English" cards.  If you are starting out maybe try to actively learn just the 625 "most-valuable" word list from Fluent Forever.  Or if you have studied Chinese for a while maybe write something on lang-8 and try to actively learn any words you couldn't produce from memory when you needed them (and hence had to look up) that you think you should be able to actively produce on demand in the future.

2. To combat interference each Meaning -> Chinese card should list all the Chinese synonyms and if you remember any one of them then you should mark it remembered ( https://www.supermemo.com/en/articles/20rules ).  Avoid trying to memorize entire sets of synonyms nor don't try to remember multiple cards for a given English word (which given the large number of synonyms is a risk if you just automatically create "reverse" cards).  If you provide additional context you could (if you really wanted to) have multiple cards testing different meanings for a given English word (i.e. "the fire was [hot]" and "calling a girl [hot] might not be the smoothest flirting strategy".

[Lumbering Ox]

5 hours ago, trevorld said:

1.  It is more efficient to acquire a larger passive vocabulary then active vocabulary so the number of Meaning -> Chinese cards (i.e. active cards) should optimally be much smaller then the number of "Chinese -> English" cards.  If you are starting out maybe try to actively learn just the 625 "most-valuable" word list from Fluent Forever.  Or if you have studied Chinese for a while maybe write something on lang-8 and try to actively learn any words you couldn't produce from memory when you needed them (and hence had to look up) that you think you should be able to actively produce on demand in the future.

2. To combat interference each Meaning -> Chinese card should list all the Chinese synonyms and if you remember any one of them then you should mark it remembered ( https://www.supermemo.com/en/articles/20rules ).  Avoid trying to memorize entire sets of synonyms nor don't try to remember multiple cards for a given English word (which given the large number of synonyms is a risk if you just automatically create "reverse" cards).  If you provide additional context you could (if you really wanted to) have multiple cards testing different meanings for a given English word (i.e. "the fire was [hot]" and "calling a girl [hot] might not be the smoothest flirting strategy".

 

 

So if I had E to C and put in the E section something like I donno, {Wombat, not [Synonym Hanzi 1], not [Synonym Hanzi 2]} that would work fine?

 

I'd much rather do E to C as I figure my passive vocab will eventually get enough work with passive reading and listening. I am more concerned with getting a solid active vocab base of textbook, chinesepod,HSK words and my opportunities to speak to people [don't know any] or writing [meh, just want to be able to pass the HSK writing test and know enough to impress chicks ;)] are limited. I figure E to C would be a good way to bump the active vocab a bit or at least give it a leg up.

 

Or should I say frig it and do C to E, is there any actual difference in building active or passive with either direction?

 

Also any point if doing either direction to write it out every time, Takes much much longer but from what I understand writing them out helps pound stuff in the fresh tasty brains. I will definitely do RTH Keyword to Hanzi with writing everything out before I see the answer so I should be good with writing out the building blocks of words.

[eddyf]

I've been using E to C cards all along. At this point I've gotten up to around 5000 words. The way I do them is using Pleco flashcards. The front of the card has the full dictionary entry including example sentences. And I have to input the Chinese word using the standard pinyin based input system on my phone (type in the pinyin then select the right characters). Most of the time since I have the full dictionary entry and the example sentences, I'm able to distinguish between near synonyms. In a few cases where two words are really similar in meaning and easy to mix up, I just let it slide and mark it as correct. This system has worked well for me, although it's certainly more time consuming than C to E cards would be. I think it's important to drill active recall and I don't really know of any more efficient way than E to C cards.

[werewitt]

1 hour ago, Lumbering Ox said:

Or should I say frig it and do C to E, is there any actual difference in building active or passive with either direction?

That's what I'm trying to figure out. I've paused all production cards recently (~2k), although I want to improve in all areas, not just reading. Funnily, this is where "sentences" might come in useful, in a sense that it is easier (disambiguation-wise) and maybe more useful to learn to produce a specific Chinese word given the English/Chinese context. Then again, prepping such cards is 太麻烦了. A lot of speaking/writing might be the only good way, to with Anki is a nerdy crutch (beyond, say, top 500 words).

[imron]

4 hours ago, Lumbering Ox said:

Or should I say frig it and do C to E, is there any actual difference in building active or passive with either direction?

I think you should do C to E.  Here's the reason why.

 

First think about what you are hoping to achieve, which from your post, you mentioned you hoped to be able to improve your passive reading and listening skills.

 

Now think about the process required for reading or listening:

 

You encounter a word in Chinese -> you process its meaning (maybe translating to English) -> you process the context -> you (hopefully) understand it -> Repeat for the next word in the sentence.

 

Nothing in that process involves seeing/hearing a word in English, translating it to Chinese and then understanding it.

 

By only training E->C cards you are training a skill that you never need to use to do the thing that you are trying to do, and while it might work out for you eventually it will be a much longer, slower path.

 

3 hours ago, eddyf said:

I think it's important to drill active recall and I don't really know of any more efficient way than E to C cards.

I do pinyin only on the front.  Word and Chinese definition on the back.

 

For this to work well, you need to make sure that you are mostly using words with 2 or more characters otherwise you'll get too many ambiguous cards.  If you come across a single character word that you want to add, then instead of adding it, find a 2 character word that embodies the same meaning (this is usually not so hard to find) and add that instead.  This will minimise cards with the exact same pronunciation.

 

As to why this is better than E->C, it's because it (mostly) meets the criteria I listed above about training skills that are required by what you want to do.  For example, going from pronunciation to word+meaning is a core skill required by listening.  I say 'mostly' because it's still not the exact skill, and if listening was what I really wanted to focus on I would make the front card audio only.  Luckily with Pleco I can switch between the two at the press of a button.

[werewitt]

3 hours ago, imron said:

I do pinyin only on the front.  Word and Chinese definition on the back.

Curious, why pinyin over just sound? Chinese Support plugin pulls from Baidu TTS, among others, and I find that TTS very palatable.

My cards differ only on the front side, back is the same for all - a big dump of all the card info, with some nice formatting and links for further study. Mnemonics and etymology for single char decks, MDBG and sentences for words.

[imron]

Pinyin is easier to test in a noisy environment (I don't like using earphones) such as on a bus/train or in an environment where other people don't want you to make a noise (at home while other people are watching TV or something).  Plus with pinyin, I can accurately recreate the audio in my mind anyway.

 

12 minutes ago, werewitt said:

Chinese Support plugin pulls from Baidu TTS, among others, and I find that TTS very palatable.

Pleco has an extended audio module with > 30,000 recorded words, and it uses TTS for any gaps.  It also has better dictionaries than the Chinese Support plugin for Anki, plus you can change the front/back of the cards at the press of a button (or two).

 

12 minutes ago, werewitt said:

My cards differ only on the front side, back is the same for all - a big dump of all the card info

That's more or less the same for me.  I have three profiles in Pleco - one for newly cards that just shows everything (I use this for cementing my knowledge of a word after it's been newly added), one with the characters only on the front and everything on the back (that I use to drill recognition) and one with pinyin only on the front and everything on the back (that I use to drill production).  Each profile has a separate scorefile, and although there's some interference (e.g. seeing a card in one profile will make it easier to know the card in the other profile) that doesn't really bother me because I'm more interested in using flashcards for short-term drilling and then using regular reading to aid with long-term retention.

[werewitt]

9 minutes ago, imron said:

Pleco has an extended audio module with > 30,000 recorded words, and it uses TTS for any gaps.  It also has better dictionaries than the Chinese Support plugin for Anki, plus you can change the front/back of the cards at the press of a button (or two).

Personal preference  I never got into Pleco flashcards, not sure why - I do try once a year though. I'm an Anki fan  Maybe because I prefer flashcards on my laptop, not just on my phone.

Re dictionaries - I believe the plugin simply polls the system dictionary (or it uses CC-CEDICT, not sure), yet I usually edit definitions based on a textbook I pulled words from. I agree there's much more info in Pleco, yet I feel it will be a case of "too much" for my purposes.

[imron]

18 minutes ago, werewitt said:

Maybe because I prefer flashcards on my laptop, not just on my phone.

I used to think this way, but then realised that wherever I have my laptop I also have my phone so it's really just as easy to use that.  It also neatly solves (or rather sidesteps) the problem of syncing between devices because I never need to do it.  The only thing where a laptop is easier is for the creation of custom cards, but with Pleco I don't need to bother with that either.  It's true that I lose out on some flexibility, but for me that's more than made up for by the time savings I get from never needing to manually create or manage cards.

[Lumbering Ox]

4 hours ago, imron said:

First think about what you are hoping to achieve, which from your post, you mentioned you hoped to be able to improve your passive reading and listening skills.

 

Now think about the process required for reading or listening:

 

You encounter a word in Chinese -> you process its meaning (maybe translating to English) -> you process the context -> you (hopefully) understand it -> Repeat for the next word in the sentence.

 

Nothing in that process involves seeing/hearing a word in English, translating it to Chinese and then understanding it.

 

By only training E->C cards you are training a skill that you never need to use to do the thing that you are trying to do, and while it might work out for you eventually it will be a much longer, slower path

 

From what I recall Hesieg discourages Kanji to Keyword. He claims that going Keyword to Kanji will develop your passive abilities whereas the reverse isn't as true. I know that it does develop the passive, it might be sluggish and you are running stories in your mind before you come up with the key word but I was nailing about the same percentage when running passive tests by looking at the lists in the back as I was in Anki with active style.

 

The question for me is, does learning vocab in the passive way lead to active as some studies have suggested, or not. Or perhaps it does, enough.

 

Actually my goals are a bit muddled.

I'd like to pass the HSK and assuming I can do it near by the Taiwan equivalent. It would be nice to have something to show for my efforts, and I don't want to be a Benny. Also it provides a path of sorts

2: I currently would have to hire out a talking partner as where I live is rather Chinese deficient aside from HS exchange students and I am almost 50 so... errr ick, awkward. On the other hand in the longer term I don't see myself being here forever, and I'd like to be able to carry on a conversation beyond "I like cats" "Where is the library" "I also think Justin Beiber is history's greatest monster" but not to the point of being able to discuss why my Marxist feminist dialectic doesn't bring all the boys to the yard and how that is a good thing. Maybe in the far future I'll move to a Chinese city like Taipei or Markham. OTOH from what I understand if you have your listening and vocab skills up to snuff, and a bit of convo ability you can get up to speed somewhat quickly.

 

I remember growing up anglo in Quebec that I never had the vocab to really use the language, couldn't think of enough words to full a thought, it's annoying. When reading or writing you can look stuff up, with listening you can just ignore what you don't know [or look it up later] but when speaking you need that word then and there.

3: Consuming audio media. Chinese movies are a mixed bag and I hear meh things about TV so for me it would be about an enjoyable way to practice instead of a driving goal on it's own. I am sure I can find something worthy. I'd like to have enough passive that I can go full native for maintenance without wanting to stab myself in the eye with a very hot french fry.

4: Reading. Again as point 3, more about the developing skill and vocab and keeping in practice. I have no particular drive in this area although I don't mind reading.

5: Writing. Once I pass HSK 6 and the Taiwan version It is a skill I'd like to keep because characters are cool and I suppose it might have some use on line but I hardly ever write in English so meh.

 

I figure the test lists, texts and pod list combined would yield perhaps 14 to 15K words without duplicates. That would seem to be enough to carry on a decent conversation. Until my situation changes, conversation practice aside from talking to myself, unless I find someone local will be expensive and limited, maybe I can swing 100 or 200 hours of tutorial. Thus my interest in Active vs passive. Everything else I can run through, texts, pods, Chairman Bao etc.

Writing, I have little need beyond the tests and any word I need in excess I can look up.

 

 

I know I am being an argumentative twit here, but this is an area where I want to be really sure I pick what is right for me. The rest of the path is pretty clear.

 

If after all the above you are still thinking C to E is the way to go to get to where I want to be, I will most likely bow down to the collective experience and follow along.

 

[werewitt]

1 hour ago, Lumbering Ox said:

I currently would have to hire out a talking partner as where I live is rather Chinese deficient aside from HS exchange students and I am almost 50 so... errr ick, awkward

Nothing awkward about it. Assuming you study Chinese with these students. If you feel awkward, hire on iTalki. Besides not giving a damn, tons of "middle aged" Chinese teachers/tutors with qualifications.

[imron]

5 minutes ago, Lumbering Ox said:

He claims that going Keyword to Kanji will develop your passive abilities whereas the reverse isn't as true

Yes it does develop your passive abilities, and reading is a passive skill.  It's also important to train active abilities, and I mentioned what I do for this above.  I think both are necessary.

 

8 minutes ago, Lumbering Ox said:

but I was nailing about the same percentage when running passive tests by looking at the lists in the back as I was in Anki with active style.

The thing to keep in mind is that Anki (or other flashcard) reviews are not really a good indicator of how well you'll understand actual content.

 

The only real test is trying to read/listen to something and seeing how you do.  If you don't do so well the answer is probably not to go back to flashcards until you 'improve' but rather to try easier content.  It's best to use content just at or outside your current ability.

 

4 hours ago, Lumbering Ox said:

If after all the above you are still thinking C to E is the way to go to get to where I want to be

Different people have different things that work for them, so while I think C to E is better than E to C for improving reading it's something you should test out and decide for yourself if it works.

 

Think about where you want to be.

 

Think about your study methods and see if they are helping you get closer to that goal or do they just involve spinning wheels to watch counters go up without really bringing you closer.

 

If the latter think about what you need to change.  Don't be afraid to abandon study methods that aren't working and try new things that seem interesting.

 

Constantly test against the skills you want to be able to use (rather than just flashcards) to check progress and also find out where your current shortcomings are.  That will give you concrete feedback about what you need to improve.

 

Periodically come back to assess and reflect upon your study methods using the above process.

 

Having said that, actually, I don't think C to E is that great - ideally you want to get away from English entirely.  I prefer C->C but that's obviously difficult to do in the beginning.  If you need to do C->E what you should do when you see the Chinese on the front is to think of the object/action itself rather than the English.  Then when you see the English that will also make you think of the object/action and you can see if they match.

 

For example if the front of the card is 筷子 you shouldn't think 'chopsticks' instead you should think this.  Then when you see the answer you can check if it was the same thing you thought of.  Obviously this isn't going to work for everything (abstract things like 而且 will be difficult) but it works for enough things that it can still be effective.  The idea is to remove as many mental processes as possible for you to get to the meaning so you can go straight from the word to the meaning without needing to walk several steps in your mind to get there, otherwise you'll never be able to process things at the speed required for native content.

 

[trevorld]

16 hours ago, Lumbering Ox said:

So if I had E to C and put in the E section something like I donno, {Wombat, not [Synonym Hanzi 1], not [Synonym Hanzi 2]} that would work fine?

 

No!  That requires you to memorize sets of info which is very inefficient.  What you want (for efficient active recall) is something like Monday -> 周一、星期一、礼拜一  (and if you remember any of the synonyms click good).  What you do not want is three different English cards such as Monday (not 星期一 nor 礼拜一) -> 周一 .  As long as you can recall (a somewhat frequent) way to get across your meaning then you are good especially if you just want to effectively communicate (and not get A's on written essays).  At some point you'll want to passively recognize the three different Chinese words for Monday but you only need to actively produce one of them.  

 

16 hours ago, Lumbering Ox said:

Or should I say frig it and do C to E, is there any actual difference in building active or passive with either direction?

 

Since there are many Chinese words you want to passively recognize but you only really need actively use much less of them you get more bang for your study time if you learn (either via extensive reading or flashcards) more passive vocab then active vocab.   Especially since if you passively consume a lot of Chinese media some of the vocabulary and grammar will become active.  Aggressively building a large passive vocabulary lets you read books, watch media, play Chinese video games, etc sooner.  Most people find they need some kind of active study crutch to get to the point where they can "extensively read" (i.e. at the efficient

98% comprehension level) Chinese media (although graded material helps).  However you do of course need separately practice writing and speaking to get good at active recall.  A crutch set of flashcards of high frequency words that can express a large set of ideas could be useful (i.e. Fluent Forever's list of 625 "most-useful" words) as well as drilling any words you  tried and want to actively use but couldn't.

 

16 hours ago, Lumbering Ox said:

I am more concerned with getting a solid active vocab base of textbook, chinesepod,HSK words and my opportunities to speak to people [don't know any] or writing [meh, just want to be able to pass the HSK writing test and know enough to impress chicks ;)] are limited.

 

http://lang-8.com/ is a terrific site to practice your Chinese writing skills, in my experience I'll write a few sentences and within a day several native Chinese speakers offer me suggestions and corrections.  Warning: you are expected to reciprocate and help out people working on their English.  If you force yourself to write without looking at a dictionary or flashcards it is a great way to practice active Chinese (i.e. try to be creative and find ways to describe words, concepts in round-about ways if you don't remember the exact right word).  

 

There are also tons of Chinese willing to practice speaking with you via Skype (either as a paid tutor or as a free language exchange).  Haven't used it but https://www.italki.com/ has a reasonably good reputation for both.

[Lumbering Ox]

On 2017-05-30 at 2:59 PM, trevorld said:

Stuff

From the totality of what you say, it seems to present a good argument for C to E. With perhaps a very small deck of E to C for important words that are not sinking in such as the ones needed for the HSKK and the written form of the higher levels of HSK. I get the impression that if one is in a position to use the language often, one can get better pretty quickly if one has a base and the passive. No reason to go nuts on it till I need it.

 

So C to E it is unless someone can blow away the collective consensus.

I am now comfortable with this.

 

 

I'll say though, when I was doing RTK I had a few words like Monday (not 星期一 nor 礼拜一) -> 周一 and found I my issue was more in remembering which one was which. I could remember all three options. However with RTK it wasn't so much Monday =3 options but rather, Monday = option 1, 1st day of the work week = option 2 and so forth as for RTK purposes it was important to map 1 word to 1 Kanji. I would guess such 1 to 1 mapping is not important for general vocab [salute the general]

 

 

 

I know of lang 8. I am not a huge fan of writing although I can get verbose at times but it is on my list of things I will force myself to do to get better.

 

I know of italki. I am a bit concerned about using language exchange. I hear it can be difficult to find a good language partner for various reasons. Also my ability to carry on an extended conversation even in my native English is a bit variable, sometimes if I can get cues I am off the the races, at other times I completely blank out and can't remember any of the awesome I've encountered recently.

Paying for someone via Italki, someone who isn't afraid to lay the smackdown on bad Chinese and draw out vocab expanding conversation seems worth it even if I can't afford to do it every day. A lot cheaper to get someone in China than Montreal, holy bejasus they get pricey.