Starting over. Your experience with LingQ?

Open Forum
41

@aronald / xxdb

Dimension 3: Efficacy <> efficiency
“BUT if you keep doing it every day till you’re complete AND you don’t care if it takes years THEN it doesn’t really matter.” (xxdb)

For the sake of clarity, we should distinguish two key aspects in this context:

  • efficacy (or effectiveness): To which degree can a specific goal be achieved or not at all?
  • efficiency: How do we use our resources (mental / physical energy, time, money, etc.) to achieve a specific goal?

Example:
Vocabulary lists with single and decontextualized word pairs (engl. chair = fr. chaise, engl. mother = fr. mère, etc.) are highly efficient to learn.

At the same time, they are highly ineffective and should be avoided at all costs (except for the absolute beginner level), as they lead language learners to produce L2 sentences that are semantically off.

“BUT if you keep doing it every day till you’re complete…” (xxdb)
Learners could study lists of decontextualized single word pairs for years (and many students at school actually do exactly that - to this day!) - just to make the experience that their speaking and writing in the L2 is completely f… ed up!

So the first question should never be: Is the approach/strategy efficient in terms of time, etc., but is it effective? Ineffective but highly efficient approaches are worthless.

The second question is: How efficient is an approach?
This question is also relevant because learners are often busy (yes, even teenagers!) and have other things going on in their life. Therefore, an approach should be as efficient as possible in the long run.

*** Reg. Duolingo in general ***
It´s absolutely ok to play around with this gamified app for a few hours at the beginning or from time to time.
However, it´s a really bad idea to do this exclusively and for months / years on end!
In this case, it is simply a waste of precious lifetime.

To take my example from a concurrent thread:

  • Learner A uses Duolingo for 25 min a day 7 days a week for 2 years. After that time he / she has an extremely high score on Duolingo, but barely understands simple sentences in a conversation, doesn´t understand the news, podcasts, audiobooks, isn´t able to read articles, books, etc.
  • Learner B uses the ultra-reading-while-listening approach (or sth. similar) for 25 min a day for 2 years (reading more than 3 million words and listening for ca. 350 hours). Learner B has reached an advanced level in reading and listening comprehension. That is: He / she can listen to podcasts / audiobooks, read non-fiction books with ease, etc.

In this scenario, Duolingo is more or less worthless for improving listening and reading comprehension and should therefore be avoided most of the learning time.

The problem here isn´t the inefficiency, but the low degree of efficacy compared to high-volume reading / listening / reading-while-listening, esp. when there are no “Duolingo stories / podcasts” available in the L2 to be studied.

Have a nice weekend
Peter

42

Hi, Davide!

Here are some other quick tips:

  1. In the first weeks I preferred a normal speed of 1x in Portuguese. Nowadays, my audio speed is 1.25x on Youtube / 1.2x on Audible.
    Only in Spanish (my much stronger language) I often prefer higher audio speeds: 1.3x, 1.4x, and 1.5x.

  2. At first, it´s better to use (popular) non-fiction like Harari´s trilogy. Popular non-fiction texts are simply easier to understand than popular fiction.

  3. It´s also a good idea to read text in German you already know from Italian, English, or French because the familiarity with the topic reduces the mental strain.

  4. Don´t do too many Pomodoro blocks a day. It´s really better to do less at first, e.g. one block à 25 min for the first 14 days.

Just give your mind / brain some time to adjust.
I mean, “ultrareading” is still a marathon, not a sprint, if you do it for a whole year - even if it’s a marathon consisting of daily sprints :slight_smile:

Good luck
Peter

43

@Davide

“Habits is not enough.”
Well, habits are only one tool. But there is also a bunch of other tools in the skill acquisition toolbox (e.g., a strong “why”, “visualization techniques”, goal setting and operating beyond goals, etc.).

“The problem comes before. Habits is an effective tool but what’s pushing a human being to unlock that will to create an habit.”
That´s a really excellent point!
We know from social psychology/sociology that the individual is not really the sovereign master of his fate, but only a node in a social network of other nodes (at least that´s one perspective, but there are also more complicated ones).

This means that our “social milieu” (i.e. our family members, relatives, peers, friends, and acquaintainces) influence our tastes, choices, learning styles, media consumption, etc. as individuals a lot!

So if teenagers want to become professional athletes, for example, they need to surround themselves with like-minded people. Ultimately, that may even mean they have to replace their friends completely. Otherwise, their old, unathletic friends will increase the likelihood that they will fail.

Dito if someone wants to become a startup founder / entrepreneur …

In short, the maxim is:
Surround yourself with the people who either already are what you want to become or share your dream. If you don t, you´ll probably fail…

For some background on these socio-psychological mechanisms, see, for example, the concept of “transclasse.”

“But usually comes from crisis, difficult situations, no choices, big mistakes, or other predispositions.”
Yes, if people want to “unlock their full potential” (let´s say, a child from a working class background wants to work for SpaceX as an engineer), they can´t be as happy as a clam.
They must be somehow unhappy with their clam existence before they can turn into someone else - no, not “Arrow” or “the Dark Knight”, but a SpaceX engineer :slight_smile:

In contrast, improving one´s language skills is usually a rather mundane thing, and even happy clams should be able to adopt a habit-based learning style.

Have a nice weekend

44

Peter,
With your ultra-listening-reading approach do you recommend the same book again but only reading it extensively without audio like a normal reading? The only disadvantage is, it is slow and time consuming at the same time I may well be doing the ultra listening-reading with a different book. However, repetition is the mother of all learning therefore reading the same book again might be beneficial in terms of picking up vocabulary and developing a natural intuition for grammar structures. Or do you think keep getting reading input from different sources(newspapers, magazines, children’s books and easy readers) in conjunction with the ultra-listening-reading approach for fiction/non fiction books?

Thanks

45

+1 @asad

46

While the Anki debt can be daunting I would highly recommend pushing through and continuing to use it. I took some time off a few months back and came back to around 1100 overdue on anki for writing and reading respectively so I feel your struggle haha. I haven’t missed a day since then though.
lingq is a great tool that I’ve been using for a long time now but Anki is indispensable in my opinion.
頑張れよ！

47

I agree on the necessary ramp up time.

My observation (from my own successful attempts) is that there is some ramp up time absolutely critically required before lingQ or watching youtube or whatnot is even slightly useful. Without it for me, it’s a waste of time.

I am not an expert by any means but this is my 4th language.

Specifically on vocabulary acquisition for a distant L2 I have a theory/tip/hypotheses:

For those who don’t like to read - skip to TLDR for the tip…

From the vocabulary accumulation perspective it’s just a raw meatgrinder if the language is far from your L1.

Unlike in Spanish/French which are not so far away from English that I can’t recognize what I’m hearing, Russian was composed of messed up combinations of consonants. To make it worse, the consonants aren’t even exactly the same as English consonants. They’re messed up wierd versions with the tongue arched in the middle and the tip of the tongue in the wrong place from an English speaker’s perspective. Not only that but there are pairs of consonants that are subtly different but which you can’t hear the difference between them.
A Russian speaker can hear the difference however.

The upshot for me was that my ability to rapidly learn new vocabulary was severely degraded. In comparison I could easily manage 50-100 new words of vocabulary per day in both French and Spanish, with the ability to reach understanding breakthrough (2-3,000 words) within about six weeks. In Russian it was not possible to go faster than 20 words per day and often that was too fast.

Interestingly, however, 7 or 8 months in I seem to have absorbed the sound patterns and clusters in Russian. They no longer seem wierd or messed up. They are somehow “normalized”. What that means is my vocabulary acquisition speed has leapt up. I reckon I could probably do double the speed now. (I’m loathe to try though hahahaha because I don’t want a massive backlog of reviews).

TLDR
Anyhow… My hypothesis is that something like Pimsleur or Michel Thomas would be epic as a starting point for a distant L2. Reason being Pimsleur breaks the words down into teeny components. This in theory should expose you to the subcomponents and burn them in.

Why does that matter?
Well if you already have the sound clusters and familiarity with wierd consonants/vowels already burned in then the efficiency of learning vocabulary should be increased and therefore it shouldb’t be such a slog to accumulate the first 2-3,000 necessary vocab words that are required for first-line comprehension breakthrough in listening.

For a practical approach: For me, my next attempt (which will be mandarin) I now plan to modify my approach from anki 2-3,000 high frequency audio words combined with TPRS youtube then lingQ to the new approach:
Pimsleur plus TPRS first THEN anki 2-3,000 words THEN lingQ.
That way, if my theory is right I should be able to get through the 2-3,000 baseline words much faster.

For comparison, I needed about 6 weeks - 2 months at the most in the case of both Spanish and French to get to 2-3,000 words whereas I needed a good 6 months to get there in Russian.
Mandarin sounds are even more wierd than Russian so some adaptation will be required.

Anyhow that’s my theory. I’m sure Peter will tear it apart.

48

As we´re all in the same time zone:

Good morning, Asad / Davide,

Yes, re-reading the same book can be beneficial, for example as “intensive” reading . But I do this very rarely. And when I do it, I re-read only selected text passages as skimming or intensive short reading.
In contrast, I usually prefer re-listening because I can do it in my down time while doing other mindless things (ironing the clothes, etc.).

Example:

  • In the morning: ultrareading session with 2 Pomodoro blocks à 25 min each
  • In the afternoon / evening of the same day or the following day: re-listening to the same text passages for ca. 30 min.
  • Alternatively as a kind of “batch” strategy: re-listening to several text passages for several hours while doing some household chores, etc.

“Or do you think keep getting reading input from different sources”
Having a varied reading / listening diet is often the recommended way in language learning. And this is a good strategy, because it allows learners to get a feel for the peculiarities of different text genres (i.e. the language structures, degrees of informality / formality, etc.).

However, going over the same material (in my case: Harari´s trilogy, Lord of the Rings / The Hobbit, etc.) again and again has also its benefits.

The question is: When do you get tired of it?

Ultimately, each and every learner must find his or her own balance between “novelty” and “redundancy”. What that balance looks like is very personal.

The only component that doesn´t change is the digestion of interesting, even fascinating “high volume” content (-> “accelerated mass immersion”), but it matters less whether the digestion process is more focused on “novelty” or “redundancy”…

Just experiment to see what works for you in the long run.

Hope that answers your question!

Have a nice day
Peter

49

“I’m sure Peter will tear it apart.”
Ha, ha, ha. You´re already negatively conditioned, so to speak :wink:

It’s an experiment so go for it and please keep us in the loop!

Personally, I´d modify your approach a bit in order to solve different problems:

  • Unfamiliar grammar structures? → Michel Thomas

  • Unfamiliar sound patterns? → Pimsleur

  • Warm up for reading / listening comprehension so that the transition to audio reader software such as LingQ is smoother? → Assimil

Esp. for a language like Mandarin, I´d use all three approaches in combo, but not necessarily at the same time (probably: Thomas + Pimsleur first and Assimil later).

After that I´d switch to LingQ / Anki.
But that´s just me…

Good luck with your experiment
Peter

50

Hi, Heinzz!

“Anki is indispensable”
Only if you don´t practice a “high volume / mass immersion” reading/listening approach such as “ultrareading-while-listening” that we´re discussing in this thread.
This works as a “natural SRS” and should be sufficient in itself.

However, I´m not against Anki. It’s a great tool, and I prefer a sort of “best of both worlds” approach by combining LingQ with Anki. But that´s just me. Toby [@noxialisrex], for example, might disagree.

Have a nice day
Peter

51

@Peter. That helps. Very interesting. Striking a right balance is the order of the day, so to speak. Just a side question, in order to derive benefits from the ultra L-R approach, if a learner is looking at a 6 month time frame technically how many books should they aim for in total? 4 books a month and 24 books in 6 months?

52

How would you define “high volume” Peter? I would argue “high volume” isn’t necessary to benefit in input based techniques (reading and listening) over SRS. In fact I would argue for those with less time, input based is better due to backlog of “review” of flashcards which means one who has little time to spend with the language will be reviewing constantly and never adding new cards/words.

I would also argue that I’m not “high volume” as I have probably averaged 5-30 min reading a day over my language learning “career” and I have still progressed nicely…albeit at a slower pace than someone who has an hour or several hours a day.

One reason I got away from SRS is because I was spending all of my limited time catching up on review…nearly 100% of my time sometimes was spent reviewing words.

However, maybe I’m misunderstanding your definition of “high volume” and maybe you mean high percentage (i.e. primarily using input techniques compared to other techniques )

53

Asad, I wouldn´t aim for a specific number of books or words read / listened to in a specific period of time.

I believe that if learners focus simply on the number of Pomodoro blocks they are able to complete each day with

  • good focused attention and
  • an elevated audio speed (between 1.0x and 1.5x),
    the rest will follow, i.e., the listening / reading comprehension, the size of the vocabulary, the identification of implicit grammar structures, etc.

Or to put it differently: it´s not a good idea to rush through books without any kind of enjoyment and sense of quality.
On the contrary, it´s better to practice accelerated mass immersion while still enjoying the process.

However, if you want more precise stats then let´s do some rough calculations:

  • 2 Pomodoro blocks a day à 25 min = on average 9k words read / listened to
  • If you do ultra-R+L 6x a week = ca. 54k words read / listened to in one week
  • That´s ca 1.4 million words read / listened to in a period of 6 months.

My Stephen King novel “It” (in the Spanish version) contains ca. 470k words according to ReadLang. Thus, I could read ca. three longer novels that have the word size of “It” in 6 months.

IMO, that´s still too complicated. The simpler solution is: just focus on 2 Pomodoro blocks a day and you´ll be fine!

Note: If you feel you can handle more blocks per day (with breaks in between), increase the daily number to 3, 4, etc.
However, your rhythm should be sustainable over time, and then “less but often” is better than “a lot but rarely”.

54

I also find that listening to native language at 1.5-2x speed helps

55

@aronald: I believe that the speed has something to do with our capability to stay focused. Which means that as soon as at normal speed our mind wanders too much, we feel the need to increase the speed to keep the same level of focus attention.

At the same time I believe that we are less focused on other things. So, probably, toggling between different level of speed could be better depending on what we want to focus on in that particular “pomodoro” slot.

56

Hi, Eric!

How would you define “high volume” Peter?
Ca. 8-10k words read / listened to in 2 Pomodoro blocks à 25 min (with a 3-4 min break between them).

If learners maintain this rhythm 6 days a week, they will absorb ca. 48-60k words in one week.

If they maintain this rhythm for one year, they´ll read / listen to ca. 2.5 - 3.1 million words. After 12 months, they should be on a B2-C1 / C1 level in reading and listening comprehension for not too distant L2s (that´s the hypothesis).

input based is better
Maybe there´s a misunderstanding here: Ultrareading-while-listening is the “input based” method par excellence. So if someone wants to discard SRS altogether, I´d say go for it.

In contrast, I´d like to test my reading / listening comprehension by doing a few flashcards (let´s say 4-6 min) right before my first Pomodoro block on ReadLang begins. Works like a charm…

progressed nicely…albeit at a slower pace
Sure, I won´t deny that. But why would someone prefer Ultra-R+L? Well, the focused attention is much better and it´s much more time-efficient (daily, weekly, monthly, and yearly).

Thus, what learner A can achieve in ca. 3 years, learner B might achieve in ca. 1 year by

  1. increasing the number of words processed per minute
  2. combining reading and listening

In short, it´s a kind of accelerated mass immersion program, esp. when learners can increase the number of daily Pomodoro blocks and maintain this rhythm over an extended period of time.

Perhaps the “shortest” answer to the question of why we should use Ultra-R+L is: because we can :slight_smile:

57

“listening to native language at 1.5-2x speed helps” (@aronald)
Absolutely!
However, reading while listening at an audio speed of > 1.5x is pretty hard.
Therefore, I feel most at ease in Spanish with a speed of 1.5x and in Portuguese (my weaker language) with 1.25x.

In general (as Davide suggested), it´s a good idea to experiment with various speeds and adjust them to our needs.

58

For anyone wondering, I listen currently to German, Swedish and Norwegian between 1,4 - 1.5x speed. Depending on the content and if I am committed to paying attention to it. If I will just have it on in the background, 1 - 1,1x.

Danish is around 1,25 - 1,3x, this mostly depends on what the app allows…

Dutch/Platt/Frisian/Icelandic is at normal speed still. I do not personally recommend going lower than normal speed personally, but I do not have well thought out reasons for this. My feeling is just it is better to struggle with content a little too fast than to slow down to unnatural speeds.

59

I do not think Anki is indispensable, but definitely useful. I decided, because of this post, to give Anki another whirl taking subs from shows I have watched and trying to translate the English subs back into the OVA. We will see how it goes.

Honestly if I could simply buy well made Anki decks, I would use them, but capped somewhere between 1 - 5% of my time with the language. 5% case being Chinese/Japanese(/Korean) split to characters vs. audio cards.

I personally like Drops and Memrise as alternates to Anki, but Anki’s customization is impossible to compete with. The problem is simply building good decks of cards.

SRS never stops being useful, but how it is useful changes through the journey. Now for recognition, I only use SRS for food, animals and technical language that simply does not happen often in writing. (Though I finally succeeded in buying cookbooks not in English.) What I am testing now is SRS for recall, lite grammar and sentence production.

60

“The problem is simply building good decks of cards.”
That´s the key - and it´s much more complicated than many people think it is!

Usually, I simply prefer the export “LingQ → Anki”. And it´s nice that LingQ not only creates words / sentences, but also contains some context info, which makes “active recall” more effective.