Improving your listening: state of the art approaches?

[Uncle Roger]

Hello!

I finally got round to learning how to use "Subs to SRS", got hold of my Norwegian DVDs, movies and TV series and all and created some 2000 Anki cards that just throw an audio string at me and reveal the Norwegian subtitle as an answer.

I'm currently a solid B2 with really good grammar and speaking, but listening has always been lagging behind (haven't lived in Norway for a while now).

I've just begun with these audio Anki content, but I can't really think of a better way of practicing listening, just as I don't think there's anything better than spaced repetition to gain the 3000-5000 headwords of vocabulary to start claiming fluency...

Or am I wrong?

I've tried with podcasts and radio but they are usually too clear and nicely spoken. I don't quite believe in watching things with subtitles because it's often a bit gimmicky to rewind back without losing the subtitle plus I really value the "lean" nature of spaced repetition. Norwegian is a major pain as a spoken language as they really hold on to their dialects so it's actually not the complicated words that fool me, but rather the small high frequency words that get squashed and eaten up.

Is there something that I might be overlooking? 30 new cards a day should keep me busy between 30 and 45 mins I reckon...

Thanks for any feedback, especially from other users of Subs to SRS.

[jeff_lindqvist]

Have you seen this topic?
LIE to a polyglot (LIE=Listening Is Everything)

[Uncle Roger]

Yes, thanks for that, I've read it now.

I agree with the importance of listening.

Reading is usually the easiest. The written word is completely machine-made, perfect. You have all the time in the word to interpret things, triangulate and have educated guesses.
The focus on writing typical of so many language tests is also a pedantic relic from the past and increasingly useless in a world of assisted typing.
Finally, I believe you can functionally express yourself in most situations with no more than 1000 headwords, i.e. no more than 3-6 months worth of studies.

Listening is the hardest thing. It's the thing that doesn't allow you to speak to natives because you don't get enough of their side of the conversation even when your half sounds believable and advanced. In my opinion, it's not your accent or your grammar that hold you back. It's the fact that in order to understand them, you'd have to ask them to repeat themselves more slowly, more clearly and with words you are absolutely familiar with. And that you'd have to ask this off them so often that they'd lose it.
The natives can understand you despite your accent and imperfect grammar.
But you/we are continuously stunned by their speed, their vocabulary and the little details unique to many of them.

So you need to know many, many words well (3000-5000 headwords?), because you don't have the time to guess, to re-read etc. And even then, it's not enough. Spoken language is akin to italic handwriting. Each person's is at least a little bit different unless they are really making an effort (i.e. writing all uppercase?).

I definitely agree with the importance of listening and I realise I have been neglecting that. What's the quickest way to catch up?
There has to be a shorter way than the 1500-2000 hours that were mentioned. I don't believe in the sparse approach of watching and re-watching whole movies or TV series. Anything you listen to and you can understand well straight away is possibly a waste of your time but it certainly becomes a waste of your time if re-listened to. If you don't need to re-listen to it, you need NOT to re-listen to it.

So, IN MY EXPERIENCE, spaced repetition is the only thing that ensures efficient use of time (and therefore a cut down on those figures of hundreds and hundreds of hours) exactly because it will naturally make you focus on what is difficult to you. It throws at you the 100 pound bell bar 90% of the times instead of 25% or 10% of the time, keeping the low hanging fruits that teach you nothing at a minimum.

Is there anything else other than SRS with audio strings as question and their subtitle as answer? Any other content I should try to Anki?

Thanks

[lusan]

I had the same concern listening polish. I tried the radio, movies, podcast, etc. And I gave up. They were either too hard or boring like hell. Now I love my new approach. I combine L2 audio books and L1 ebooks into Anki cards. I began with 1min sound segment cards. Now I do 6 min cards. (Eventually, I plan to do full chapters. My target is to switch to all-L2 cards by year end.) So I listen Polish while I make sense of the listening by reading in English. All cards are passed in Anki. I notice listening improvements. Maybe is LR without the drag. By the way, I am using Agatha Christi' novels.

[reineke]

I just listen. More importantly, I listen to a lot of easy, fun stuff. If you listen carefully you'll find that such programs regularly introduce low frequency items.

[Sayonaroo]

I recommend listening to the sub2srs audio files separate from anki reviews. I think after you reach a point it's more efficient to re-listen to an ep of something over and over after looking up words/phrases rather than doing a sub2srs deck (it seems like your level is high enough to the point that the time it takes to make and do sub2srs decks aren't worth it imo. maybe if you super tweak your anki settings to really push them out and make the steps into 1 step that's 9999 minutes etc etc maybe it's not bad). I just find sub2srs decks time-consuming because you have to make it and listen to the audio or relisten to it and it adds up a lot when compared to decks without sound. It also depends on your format. are you going to make it audio on the front, text/translation on the back OR audio and transcript on the front, translations/dictionary look-ups in the back. I think the first format is a recipe for frustration because there's no context compared to actually watching the show (the screenshot may or may not make a difference) and should be avoided.

I do not find sub2srs decks lean and i eventually deleted mine (It just got frustrating as well from lack of context since usually you have more clues/hints to help you follow the show etc when you actually watch the show vs some isolated line in a deck). What's even more effective for listening is to have each line repeat 2 times or more like when i use mp3albummaker to join the audio files for sub2srs if I load the folder that contains the audio files twice the resulting file is has each audio file repeating 2 times before moving on to the next audio file. It's annoying but more effective than the non-twice repeating version.

My recommendation is to watch with subs, LOOK up whatever really bothers you, and then relisten to the ep multiple times since if you can't catch something it's either you don't know the words or you just can't catch it. And as others have said you just gotta listen more so I also recommend reading and listening in other ways like audiobooks, podcast, etc. You gotta put in the time

[Uncle Roger]

Thanks for the interesting replies and suggestions.

Lusan:
I see what you are doing, I'm not sure what you are trying to achieve off the 6 minute segments. To me, that type of listening exercise is a legacy from the typical scholastic approach to language learning and very detached from reality. I can't remember I really had to listen to someone for that long but again, your taste and aims might be different. I'm really after improving my understanding of what others say. I find that I have to ask them to repeat themselves more clearly to then discover that what they had said was very basic. Really frustrating and very much about simple conversations.

Reineke:
The low frequency words are hardly my problem. If anything, the more you leave behind the core, generic few thousand most used lemmas in the language, the easier it is to guess what the low freq items are, as you start to meet a lot of technical words, long words, loan words or words that are not too different from their counterparts of the languages I already know. Norwegian is a Germanic language after all and despite I don't know any German, English can help enough to guess low freq words well, even when listening.
I don't know if this is a problem typical of languages whose spoken word lends itself to more "mistreatment" and inconsistency than the average language. For instance, I find (Castilian) Spanish rather "well spoken". I have never studied it formally (unlike French) but I can understand a lot more spoken Spanish than French by now. Listening to most Spanish seems as straightforward as reading it to me.
Same for Japanese. I don't know much of it, but I can recognise quite easily the few words or expressions I know, if I hear them.

Conversely, I get the feeling that some languages allow much slack and inconsistency to their speech. I know that Danish children are the last to reach a certain understanding of their parents' speech compared to other European languages. I fear Norwegian is not far behind in this...

So, again, it's a case of me being often unable to recognise short words that are surely among the 1000 most used in the language but that get squashed between other words and mumbled and thrown casually whenever people get "lazy" or when non-RP dialects come into play.
This makes me think I really have to focus on badly/casually spoken Norwegian.

Sayonaroo:
Thanks for your detailed reply. It probably takes me 20-30 minutes to turn a movie into a deck with Subs2SRS. From what I have seen, that's anything between 500 and 1000 cards. At 30 new cards a day (already quite a high number), that would keep me busy for at least 20 days of new cards plus another 20 maybe for proper assimilation.
Daily, it's between 30 and 45 minutes to review some 80-100 cards.
I think Susb2SRS is a good return on investment and, right now, I don't think I can spare much more full attention than that on a daily basis.

However, assuming a card's audio can be some 3 seconds, I am actually listening to 5 minutes worth of content for a day in which I study/review some 80-100 cards. That's actually very little, as watching TV for those 30-45 minutes will, obviously, make me listen to at least some 20 minutes of speech (we have to assume people don't talk incessantly).

So, although I agree with you that spending those 30-45 mins a day just watching TV would give me more "mileage", I think the "quality" of what I listen is important. By having the audio deck on Anki, I can focus on anything that is hard for me and keep on scheduling as Easy anything that I get straight away (or even suspending the card).

I also agree with the extra difficulty of the lack of context in my Subs2SRS decks (I purposefully avoid the screenshot), but that's part of what I am after. If I can understand an audio string randomly thrown at me with no context, no chance of lip-reading etc, surely I'll have even more chances of getting it when everything else helps me. I'm going for a "train heavy, then perform light" approach.
This is also why I have stopped with TV series and movies. I tend to remember them too much. The more I watch them, the more (I think) I just fool myself that I can understand everything just because I can tell what's going on. Plus re-listening all the things that I was able to tell the first time around is a non-lean approach imo.
It's a bit like the growing gap between listening and reading. When you read you have more time to have educated guesses, to re-read and that makes you overestimate your vocabulary knowledge and listening skills. Watching whole TV/movies feels the same to me.

I like your idea of a playlist of media from Subs2SRS, I might try it with all the harder cards later on.

Again, my situation is probably a bit strange (or rather, we all live slightly unique learning situations), but I really feel it's a case of selectively listening and re-listening to specific instances of "non-standard pronunciation" of words that are actually quite high in frequency.
I think this need of mine is a combination of my approach so far (too much reading, not enough listening to enough sources with enough variety) and something inherently difficult with Norwegian (strong dialects, plenty of non-standard inflections, prosody, you name it).

Apologies if I seem a bit critical of many proposals, it's great to read them in any case and might be useful to many others.

Happy to continue discussing this.

[reineke]

Danish etc

The authors were investigating whether Danish is more difficult to understand for speakers of the other two languages. The results show that Danish has a significantly higher canonical articulation rate which results in more syllable deletion compared to Swedish and Norwegian, which might make Danish more difficult to understand.

We hypothesize that similar differences are found between Spanish and Portuguese...

https://forum.language-learners.org/vie ... 02&p=54218

When too many vowels impede language processing: The case of Danish

https://forum.language-learners.org/vie ... 02&p=55101

A phonetician's slant on this issue

Language learning with the radio

In 1940 Max Mangold bought his first short wave radio. He already had a good phonetic knowledge of most of the languages spoken in Europe, and with a short wave he could listen to all of them at will...

Since 1940 until today Max Mangold has listened to foreign languages over the radio every day. Some days he has listened to as many as 20 languages a day. On other days he has followed individual languages and individual broadcasters. He has concentrated on individual words and expressions or on individual pronunciations and pronunciation variations. Short wave broadcasters repeat their news very frequently and for a language learner these “old news” are very good news, because the same words and expressions are repeated over and over again.If announcers change in the course of the day, you can experience different pronunciation variants and even different dialects.

https://forum.language-learners.org/vie ... 02&p=92520

The influence of foreign scripts on the acquisition of a second language phonological contrast

Besides revealing, for the first time, that foreign written input can significantly hinder learners’ ability to reliably encode an L2 phonological contrast, this study also provides further evidence for the irrepressible hold of native orthographic rules on L2 phonological acquisition...

...attaining mastery of a confusable L2 distinction would deliver a huge gain for learners perceptually.


https://forum.language-learners.org/vie ... 02&p=55155

[Uncle Roger]

Thanks reineke... is that evidence in favour of my ramblings? Can't make perfect sense of all of it...

[reineke]

Hi

I was not trying to prove you right or wrong. You raised some interesting points about linguistic similarities, phonological distance, Danish, Norwegian etc and I thought to provide some background information.

No one has as of yet made perfect sense of the language learning processes.

"Comprehension is where language acquisition begins, but what happens after comprehension is achieved is so complex that there is not a scientist on the planet who can figure out exactly what happens during the process."

https://forum.language-learners.org/vie ... 02&p=87377

Individual learner characteristics. "While some learners appear to have blinders on, in that they do not take notice of even obvious cross-linguistic similarities, others are too prone to assuming similarities where they do not exist. "

https://forum.language-learners.org/vie ... 02&p=90871

The SRS approach could prove useful. Keep us posted. Start a log.