Learning Mandarin by watching videos

[lechuan]

She's only spoke Japanese (mother tongue).

[Demonic_Duck]

As well as what Lu said about children needing interaction in order to learn their L1, bear in mind that young children also have a vastly greater quantity of input available to them than your half-hour a day, or even your ideal situation of 2-3 hours a day.

 

Someone mentionned that babies get much more time of exposure. I would add that they also get interactions and mediations. By the TV-only method you can guess some expressions about some actions and names. But without interaction, subtitle or any kind of mediation, even with a super-memory I can't see how you could decipher abstract structures.

 

Why do I get the feeling we're going in circles here? ;)

 

@lechuan: that's an interesting case, as whilst Japanese is closer to Chinese than, say, English, it's not that close. However, it sounds to me that her method was significantly different from what the OP is doing. Firstly, you say she learned Mandarin almost exclusively from watching subtitled dramas, and reading. Well, how exclusively is "almost" exclusively? Was she living in China at the time or had a Chinese friend to talk to? With the reading, I assume this was, at least at first, with the aid of a dictionary, looking things up and gradually "decoding" sentences, which is hardly passive learning.

 

Secondly, you say that she drilled the sentences from her subtitled dramas over and over again, and I'm assuming she didn't do this with sentences she didn't understand. How did she come to understand them? I assume, again, she looked them up in a dictionary and decoded them.

 

Thirdly, you say she "easily surpassed your level", which you list as "upper beginner" on your profile. From personal experience, though I think it's easy to decently evaluate people of an obviously lower ability level than your own, it's much harder to evaluate those of a higher level, and you tend to assume that anything of a significantly higher level than yourself is basically indistinguishable from native-level. So, we really have little idea of what level she actually achieved.

 

Anyway, I'd be curious to know more details as to her exact methods and such.

[jbradfor]

 

Nobody's going to be able to check those out unless they're Prime members, btw.

Took a quick listen and it's pretty standard Mandarin. I didn't watch much (being at work now...), but offhand it doesn't seem to be a good source to learn.  Lots of time with no one speaking, some somewhat technical / obscure vocab, and with the English subtitles I'd guess you're going to pay more attention to the subtitles than to the audio.  On the plus side, he speaks pretty slowly.

[abcdefg]

Please check it out, and if somebody could confirm that it really is in standard Mandarin, I'd really appreciate it!

 

For #16 -- @victorhart -- I watched that documentary and think the narrator speaks good, standard Mandarin. It would be a better learning tool, however, if it had Chinese subtitles as well as (or in place of) the English ones.

 

Edit: I didn't see @Jbradfor's post.

[lechuan]

@Demonic_Duck,

 

It's true, they are quite different approaches. I guess the only common thread is that they both make use of videos, albeit in very different ways. I know I learn a lot more in 10 minutes of mimicking and copying portions of a video than I do passively watching 1 hours of video (Just tried it, I passively watched the first episode of iPartment and didn't really seem to learn anything. Then went through the first few minutes repeating the characters lines over and over again and learned some new vocab and sentence structures). For sure, looking up unknown words.

 

Ah, I did not think about 'immersion' when I mentioned exclusivity. We live in a sizable mandarin community, and married into the same Mandarin-speaking family, so lot's of opportunities to practice the language, if so desired.

 

Based on the ACTFL Proficiency guidelines, I would rate her in the advanced spectrum, at least for speaking and reading, at least from observation of communication ability with family/friends, and levels of reading materials.

[Demonic_Duck]

Don't get me wrong, I definitely think there's plenty of utility in passively taking in content (I don't meen "passively" as in listening to tapes in your sleep, I mean in a way that involves focused attention but no interaction). It can certainly help with general comprehension skills (highly useful in themselves), although not so much with vocab acquisition. Personally, if I find a piece of content worthy enough of my attention (main criteria: decently interesting and around the right level for me), I like to go through it a few times - first passively, for gist, then actively looking up and drilling words, and then finally once through again to check that I remembered everything. Practice general comprehension, then acquire some vocab, then reinforce that along with a higher level of general comprehension.

 

It's when we're talking about only that first stage, and especially when starting from a level of zero ability, that I think this method would be inefficient in the extreme.

[renzhe]

Just tried it, I passively watched the first episode of iPartment and didn't really seem to learn anything

iPartment is a very advanced show in terms of language. It only becomes useful for learning once you can understand most everyday spoken language without much trouble.

Also, you can't learn much from 1 hour of listening. 1000 hours, maybe.

[lechuan]

 

 

iPartment is a very advanced show in terms of language. It only becomes useful for learning once you can understand most everyday spoken language without much trouble.

Also, you can't learn much from 1 hour of listening. 1000 hours, maybe. 

 

Ah, so would videos would be similar to a graded reader situation, where it's best to choose material that we can already recognize a large percentage of?

 

Yeah, I didn't have time to try 1000 hours of passive listening So it will be interesting to see the results of this experiment

[renzhe]

Ah, so would videos would be similar to a graded reader situation, where it's best to choose material that we can already recognize a large percentage of?

Even moreso, because you can read at your own speed, whereas you typically have to follow spoken language at full speed.

[victorhart]

Thanks everyone for the posts! I will respond in more detail in coming weeks, when I have a bit more time.

 

First off, I'd like to ask you to visit my blog, www.mandarinexperiment.com, and comment there as well (including "negative" or "discouraging" comments, that's fine). It's really exciting to me that my 6-year-old daughter wrote the blog post this week, without any help from me. See what she has to say about the experiment! (The text at the top is an edit I did of what she wrote).

 

What many of you are saying is that my approach will be ineffective or at least very inefficient. Time will tell! I have put some additional thought into this in the past week or two, especially because now my daughter is doing a lot of Chinese viewing. God forbid I waste my precious daughter's time!

 

However, I am basically at peace with this. Here's why.

 

First of all, although I am open to being proved wrong and some of you make good points, I am no neophyte when it comes to language acquisiton. I speak and write English and Portuguese as native languages, my written and spoken Spanish are advanced enough that I often fool native speakers into thinking I myself am native, and I used to have advanced fluency in French as well (it is now very rusty). I have dabbled in Arabic and Swahili. Further, I have spent a lifetime observing people's various approaches to language acquisition, what works and what does not, including about 14 years doing this as founder of a language institute (where 100% of those who have taken TOEFL, IELTS, DELE etc. have gotten the mark they needed).

 

What my daughter and I are doing in Chinese is different from all that. Mostly because Chinese is an entirely unrelated language to those I speak, and secondly because, for the purpose of "science," I am being radical in my approach, i.e., not doing any formal study at all and (much worse) having no conversation or tutoring.

 

My experience in language acquisition assures me that we will learn. And in the first two months I have already begun understanding dozens of words.

 

But maybe my method is extremely inefficient, as some of the posters claim. I actually don't think it will be. I think it will be significantly more efficient than most approaches. But I grant that I could be entirely wrong.

 

However, the second reason I am at peace (at least at this point in time), is that if not for this experiment, my daughter and I would not be studying Chinese at all. We couldn't really fit it in or manage it otherwise, because of our busy lives in Brasilia. We are bonding over it, we are enjoying the shows and movies, and we are broadening our cultural understanding. Blogging is interesting and purposeful. So even if I eventually come to the conclusion that our method is very inefficient, that will be OK.

 

Just one more comment. I still haven't analyzed Keith's experiment very carefully. But from what I have seen, the conclusion that he wasted his time is far from evident. You cannot judge his progress by his first attempt at conversation. I make no claims that I will be at all conversational after my 1,200 hours of listening. I only hypothesize that I will have good oral comprehension. I would also hypothesize (but this will be a second, thus far unplanned phase of my experiment), that I will then be able to effectively and quickly (I am speaking in relative terms, of course) add speaking to my skill set by hiring a teacher or spending time in China, followed by reading and writing the script. Yes, kind of like a kid does.
 

That is my defense and a few quick reflections. But please do keep the posts coming, including skeptical and critical ones, since they are all very enriching to me and hopefuly to others as well!

[jbradfor]

Not trying to be snarky, but does the previous post remind anyone of a certain Irish polyglot we had a while back?

[victorhart]

jbradfor, tell me about him

[jbradfor]

I believe the user id was IrishMike??  He, like you, had a "New System" for learning languages (just conversation, IIRC).  He, like you, already spoke a number of languages.  He, like you, came here "for advice", but had basically already made up his mind that his system was the best, and wouldn't listen.  Not sure I should have brought this up: normally we are pretty friendly, but he rubbed a number of people the wrong way, and things got "tense".  [Actually, he rubbed pretty much everyone the wrong way, some showed it more than others.]. Ultimately, he failed in what he said he would achieve, and we never heard from him again.

 

I hope that you, unlike him, will listen. Many people here have spend decades and thousands and thousands of hours learning Chinese, and tried pretty much every way (including watching videos, but not as a beginner).  [i don't really count myself as among them.].  If you want to go ahead and try it for fun, I certainly wouldn't want to stop you.

 

I guess what I'm trying to say (maybe not in the best way) is that please don't get so caught up in trying to "prove" something that you waste your time (and your daughters), beyond what is reasonable.   Give it a reasonable shot if you feel you want to do it as an experiment, and then acknowledge defeat -- or victory, if it is working.  If it's not working, come back here and I'm sure we can find other ways to learn that are fun for you and your daughter.

 

And please use our posts about learning Chinese from watching TV series, in particular the easy ones to start.  If you do, post your questions and opinions.  And we hope to continue hearing from you, irrespective of how it turns out.

[Lu]

Don't you mean Benny? I forgot his username, but I don't think it was 'Mike'. Victor Hart is not really reminding me of him so far, he seems a lot less dismissive of other people's approaches...

 

Just one more comment. I still haven't analyzed Keith's experiment very carefully. But from what I have seen, the conclusion that he wasted his time is far from evident. You cannot judge his progress by his first attempt at conversation.

If that's the best he could do after, what was it, two years of daily 'study', then yes you absolutely can judge his progress, or lack of it. I'm somewhat impressed that he managed to get to conversational level after just passively watching tv shows, but I'm not at all impressed by the level he managed to reach. There's a guy in my Chinese conversation club who has a similar level and he has been studying for a lot shorter. If Keith had used other methods in addition to watching tv, he would have gotten much, much farther. If your position is that watching tv shows in the language you're learning is beneficial to your progress, I wholeheartedly agree. But learning by only watching said tv shows is not efficient. It will probably lead to some amount of fluency eventually, but it will take a lot longer than if you would use a mixed approach.

That said, I know you're approaching this as an experiment to see how far you'll get, and as an experiment it's perfectly valid. I hope you keep it up.

[victorhart]

Hi everyone! I've been continuing my experiment daily. So far, 47 hours of viewing in 86 days. Impossible to measure progress with any objectivity, but I feel it's going well. And I’m enjoying it! 

 

Lately I’ve been enjoying Chinese cinema. In particular, I would recommend Journey to the West (very funny and crazy), Hero, Fearless, and Crouching Tiger Hidden Dragon (all three traditional kung fu / Chinese epic type films). For those that like romance, The Road Home is high quality. And Riding Alone for Thousands of Miles is quite good, but the downside is that it is half in Japanese. 

 

Regarding Keith, I’ve exchanged emails with him and gotten his contact on Skype, but we have yet to talk. I’m very curious to hear more directly from him, and when I do, I’ll post a summary here. Despite the criticisms here and on his blog, I actually think it’s amazing he was able to have any conversation after just listening (albeit for 2,000 hours). The skills do not transfer automatically. After my 1,200 hours, I hope to have an intermediate level of oral comprehension, but zero speaking ability. I will then take on speaking as a separate skill, but I expect that my gains in listening comprehension will allow me learn to speak much more quickly than otherwise.

 

耳耳语语 and others: I agree with the importance, for children and adults, of interaction and mediation. I’m not getting any interaction or mediation in my experiment (except for the occasional subtitles). So that is one of the key issues that is being put to the test. I assume that interaction and mediation are very helpful, including for listening comprehension. However, I hypothesize they are not indispensable, and that is a large part of what I am testing.

 

Lechuan, thanks for commenting on your sister-in-law’s experience. Though I agree with Demonic_Duck and you that her experience is quite different from my experiment, it is nonetheless an interesting case and heartening. And though a lot of repetition is not part of my experiment, I already do a little of it (without pausing video), and I’m sure I’ll do more as I learn more. In fact, I’ve always noticed that “echoing” in my mind what I’m hearing when listening to a foreign language is an important part of the acquisition process, and I think I (and probably most people) do it all the time. That's different from actually pausing the listening to repeat, but I think the principle is similar.

 

Jbradfor, thanks for the information on that amazon prime source I posted. As for your comments on IrishMike or Benny, all I can say is that it is not my intent to be obnoxious or deaf to suggestions. However, it is my intent to do a serious experiment, which requires isolating variables, and being persistent, giving it adequate time (probably the 1,200 hours) to provide results. I’m not exactly trying to prove something, but rather to test a hypothesis while enjoying myself and hopefully making some progress toward a worthwhile goal.

 

As to the general efficiency criticisms, I can certainly say that if I had taken more traditional approaches and in particular used vocabulary lists, I'd have a lot more to show for the 47 hours. However, in learning other languages I've found vocabulary lists ineffective in the long term and listening to authentic sources (along with reading and speaking) much more effective. Also, please keep in mind I don't advocate a listening-only approach as the most efficient, but rather believe the most efficient approach is a combination of listening to and reading authentic sources, speaking as much as possible with natives (with feedback and corrections), as well as writing. Please remember: I'm listening exclusively in order to isolate variables.

 

But also please keep in mind the fun and convenience aspect of my project. I live in Brazil and have an incredibly busy schedule, so watching 30 minutes of video with a midnight snack is far easier than hiring a teacher or planning trips. And I'm really enjoying watching great Chinese movies, as well as Boonie Bears with my daughter! That is so much more fun than looking words up in a dictionary or memorization of any kind. 

 

I’m not being at all dismissive of other approaches. I think that should be clear since I believe my method is definitely not the best. I also think depending on people’s goals and circumstances, a wide variety of approaches can work.

 

I’m looking forward to making a table compiling Mandarin language films that will hopefully be a good resource for other learners. And also to getting lots of suggestions on high quality movies that I might not find on regular movie lists (and hopefully and how I can easily access them).  

 

Oh and if you’re interested, do visit my blog, www.mandarinexperiment.com.

[victorhart]

I'm happy to report that I'm on track with my Mandarin experiment.

 

I just completed 240 hours, or 20% of my total time, ahead of schedule. I'm having a really good time with it!

 

As I've posted on other threads, my biggest and most pleasant surprise has been the quality of many of the Chinese films I've watched.

 

I've also mentioned my two big discoveries in terms of TV shows:

 

(1) Qiao Hu, which is certainly much more entertaining for two year olds than for adults. If you can stand it, however, it is extremely useful content for beginners like me. Upon a suggestion from an administrator of these forums, I began developing Qiao Hu study guides, which evidence my progress but are also useful, I hope, for other beginning and low intermediate students.

 

(2) Boonie Bears, which I consider very useful for beginners who enjoy slapstick entertainment and just want to get a lot of Mandarin in their ears, since you can easily follow the plot without understanding the dialogue. I'm guessing this show might be even more useful for intermediate-level students.

 

What about my results?

 

Well, one way I found of tracking them, which I initially did more for fun and out of curiosity, but which I now consider actually useful, is estimating what percentage of words I understand on first viewing of an unfamiliar show. I include repeats, so I call these "word occurrences". Here's a little graph that illustrates my progress over time. 

 

 

I now understand something like 8% of word occurrences, which is unimpressive but represents significant progress for me. At 120 hours, I was at under 3%.

 

It's still too early to come to any conclusions. However, extrapolating (highly tenuous) projections from my current comprehension, and considering my initial hypotheses, I conjecture that:

 

- my 1st and main hypothesis, that I can learn Mandarin in this way, will prove true

- my 2nd hypothesis, that this method is more efficient than traditional, old school methods, may prove false

- my 3rd hypothesis, that I will attain an intermediate level of comprehension after 1,200 hours, may prove false 

 

But since I am learning and having a good time, I have every intention of seeing this experiment through to the 1,200-hour mark!

[pancake]

I think your experiment is very interesting. From the data you have collected I would worry about the limits of this method. The frequency distribution of words in natural languages follows Zipf's Law (so that, for example, learning the first 10% of all words in a language will take much less time than the next 10% and so forth). You have been at it for almost a year, and if your method worked well I would expect a much steeper slope as you are "picking the low-hanging fruit" language-wise. As it is, your overall comprehension is still low, and you have not even encountered the low-frequency stuff yet (as evidenced by your slow linear increase in comprehension)!

[Lu]

I'm impressed that after ten months, you're still at it and glad to hear you're enjoying your experiment! Interesting to see the results.

Pancake, I think that for this particular learning approach, the steeper slope will come a bit later. I think that once you hit a certain percentage of comprehension, new words suddenly become easier to learn. Understanding 8% is virtually nothing. There will be no sentence where you understand everything except one new word. Once you get to 70 or 80 percent, you'll have more of those sentences and you can learn much more from context.

But that's just my theory. We'll see how it works out for Victor!

[Silent]

 

I now understand something like 8% of word occurrences, which is unimpressive but represents significant progress for me. At 120 hours, I was at under 3%.

I understand it's just an estimate, but wonder how you come to this estimate. you're not underestimating? I would not expect that just watching would yield big results but 8% is, if based on word count, even worse then I would expect. I mean 8% is basicly just a hand full of words.

[Demonic_Duck]

Wow, yeah, looking at that wiki article on Zipf's law, which states that (based on the Brown Corpus) "the" accounts for around 7% of all English words, that really looks very rudimentary indeed, although I don't know what the statistics are for spoken Chinese. I guess you're probably also missing out on a lot of very basic function words and instead learning somewhat less common content words first.

 

Edit: based on this 1.2 billion word Chinese corpus:

 

Single most frequent word: 3.52%

Top 5: 7.92%

Top 10: 11.78%

Top 20: 17.51%

Top 50: 29.40%

 

Of course, that's for written newspaper Chinese rather than spoken. I suspect that if anything the ratios would be more skewed in favour of high-frequency words in spoken Chinese.

 

Edit 2: looking more closely at the list, it looks as if the word segmenting capabilities of whatever software they used may have been sub-par, as this is mostly a list of characters, with only a couple of multi-character words included. Not sure what effect counting true words rather than characters would have on the results. The single most frequent listed, however, is definitely a distinct word (though sometimes included as part of multi-character words, albeit with a different phonetic value).

 

的