Is extensive or intensive reading faster for vocabulary acquisition?

Really? Like they banned you based on your chat with AI about optimizations of xenomorphs? These OpenAI are so open… I’d vote against Sam Altman’s being in Congress or even close to Congress, I’m suspicious of this guy for some reason :smiley:

Lil monster is cute! And perfectly suits to the terrifying tendency of AI application. They’re selling AI based systems for even better controlling employees in their workplace, did you know about that?
AI enthusiastic people would go on and on about Neural Nets making our life easier, but in the meantime they’re going to help authorities to enslave us more effectively, like starting right now.
Or to recongnize the type of traffic we’re getting and sending to just block everything of this type without the need for details. In the Ministry of Truth they’re testing such systems for blocking VPNs, I guess it’s just the greatnessness(-nessness) has been shaken by all those dangerous leaks of foreign information :]

4 Likes

I think what really would improve the lookup problem on lingq would be a combined sentece mode. Currently LingQ has.

  • Normal mode (TL only, full page).
  • Sentence Mode (TL and Translation, but only one sentence and you have to click on the translate button)
    A great addition might be a combined sentece mode (TL and Translation under each sentence, without the requirement for additional button presses, and fill the page with as much sentences as possible).
3 Likes

EXACTLY. This would massively reduce the search. I would definitely pay to buy an internal dictionary. It wouldn’t be enough BUT it would be for 90% of the time!

3 Likes

They banned him because he was actually being effective in doing so! :laughing:

2 Likes

Nothing new to add, but thank goodness for the new message board. I would’ve given up on this thread a long time ago on the old message board as I would never have known where the new posts were hidden =)

7 Likes

Having only read maybe half of this thread, I agree with you, but have also come to the conclusion that 2.500 word threads are hard to follow.

5 Likes

Can’t you already display translations under the sentences? Although I think that’s only if you have a translation saved, not if it needs to be generated first.

1 Like

Well, OpenAI is not “open” bc. many topics are taboo (sexuality, politics, religion, violence -
and “monster optimization” :slight_smile: ).

IMO, an AI superintelligence evolving from LLM is probably nonsense. However, what is a real threat is that AIs become weaponized (beyond the military) bc. they are increasingly part of all the ongoing global conflicts (states, warlords, terrorist organizations, and organized crime).

Anyway, OpenAI probably banned me bc. I wanted to devise a presidential campaign for Trump for 2024:

  • Monster hybrid: Trump and Pennywise (looks “cute” thanks to Craiyon!)

  • Slogan: “Make America eat again!” :grinning:

  • Advantage: When a giant alien spider wants to hunt you down and digest you, it doesn’t matter which skin color you have, if you’re rich or poor, pro-Trump or against Trump, etc.
    Meat is just meat. There’s something liberatingly egalitarian about such an alpha predator logic, right?

6cbde78b8f8f46839a848f99ad09b458cce9c060

Or, maybe, an even better campaign slogan would be:
“America, let’s float together!” :upside_down_face:

4 Likes

No, I didn’t know that.
Usually, that’s an utterly stupid idea bc. the demotivation of the workforce tends to skyrocket…

2 Likes

Yeah. What a time to be alive!

Heard somewhere recently:
“Why is that everybody around seems to keep recalling 2007? Just a regular year.”
“Yeah.Exactly.”

Yeah. Exactly. :slight_smile:

1 Like

Well, I’ve changed my mind, at least this one works faster, time’s precious…and nothing else matters :]

2 Likes

Without your comment I’d not have known this was supposed to be a hybrid at all. I don’t what this tells me about the world.

“I don’t what this tells me about the world.”
Well, we don’t have access to the “world”, but constantly use all kinds of media to construct our realities (at least from a constructivist perspective). The “world” just tolerates these reality constructions or it doesn’t (then one gets hurt or dies).

Anyway, the interpretations are manifold (some may be obvious, others not)…

One maybe not so obvious background story is this:
Nowadays, I’m being bombarded with messages how much Germany, esp. the young generation, has to suffer - even without being massacred in Ukraine: post-COVID depression, economic crisis (recession), rising inflation, bla bla bla.

Me: “Really? There are a thousand ways that the “worst” day in Germany can feel like paradise.”

Nice that the YT algorithm agrees and recommends me a video of a young woman from near where I live being torn apart by a crocodile in Australia a few years ago… Well, I was innocent: I was just watching “Running Wild” with Bear Grylls :slight_smile:

Must be the influence of too many Goggins / Jacko vids in the past…

2 Likes

It comes to my mind a very interesting book: “Man’s Search for Meaning” by Viktor E. Frankl. Original title: " Ein Psychologe erlebt das Konzentrationslager". Man's Search for Meaning - Wikipedia

My grandfather survived back home from Dachau.

3 Likes

I agree: a classic!

On the other hand,I was confronted with teenagers who told me that
their math “test” (for ca. 30 min) was worse than the Holocaust.

You can imagine my reaction…

4 Likes

Oh wow, what an interesting topic! I haven’t been on the forum for a while and I’ve missed so much!

I know I’m late to the party, but I wanted to add my perspective as I’m not sure it’s quite been covered here.

My native language is English, I have learned Swedish to what I consider fluency (B2-C1) and am currently learning Turkish. This information is important later on.

I wanted to contribute to this because when learning Swedish I did not use LingQ (I found LingQ very late in my Swedish journey and by then I had my method and LingQ didn’t really suit it).

To learn Swedish I mostly used extensive reading/ listening. Mostly as in, near 100%.
Initially, when reading I did all my reading while listening to the audio at the same time. I did not stop to look up words (and on the very very rare occasion I did, I hovered over the word for a split second, nothing more). I used the audio to keep a pace, and essentially to stop me from looking up words. And then I branched out to listening without reading.
I listened to and read content that I mostly understood. I started with Harry Potter as I had read that before in English, but quickly became bored and just started reading things that looked interesting. When listening, I also never stopped to look up words. With both listening and/or reading, I could either work out the meaning from context, or I couldn’t, and I accepted that there was sometimes a sentence or sometimes even a paragraph I didn’t quite understand. Sometimes I would hear/read the word in context a few times and then realise I retrospectively understood something a few chapters back. Or I’d realise I’d understood something completely incorrectly until this point in the book. But since I was enjoying the process it didn’t really bother me. I laughed it off and continued.
I only really ever looked up words when I wanted to say something, not understand something.

For reference, I would guess I got to a high B1 level (in comprehension, not output) in around 4-6 months using this method.

But here’s the thing… Swedish is a very similar language to English. The grammar is essentially the same, and there are lots of words similar enough to English that I could guess the meaning correctly. Another major advantage is that I’m Scottish, and the number of words in Swedish that are incredibly similar to Scots words is crazy - words I suspect only Scottish people would be able to guess, and English-only speakers might not.

Then I picked up Turkish. There are basically no similar words (except ‘modern’ words related to sports or internet etc), the grammar is incredibly different, the word order… there’s nothing for me to work with. Nothing about this language is intuitive to me. I have found it utterly impossible to learn through extensive reading, and that threw me initially I’ll admit.
I’m using LingQ now (obviously!), and I’m mostly using intensive reading, still in sentence mode. I’m looking up most words. When I attempt to read a little more extensively, I’m lost. And not in the way I was with Swedish when I didn’t really understand a sentence properly, I mean lost in a whole other way.
This time round, instead of just reading and listening together and moving on, I’m reading through the content first (without listening), looking up words and getting a grip on the meaning first. Only then do I read and listen without stopping. I feel like this approach is slowly slowly working.

But to compare to my experience with learning Swedish, I’ve been at this for nine months now and would put myself at a high A2/ nearing B1 level (comprehension only, not output). Big difference, both in language ‘difficulty’ (for me) and learning method. But as far as I’m concerned it’s the language difficulty that has dictated the method.

I’m so grateful LingQ exists, and I love learning this way, but I do worry that we become dependant on it a little. Maybe that’s the extremes in learning methods I’ve followed talking; going from looking up zero words in one language to looking up every word in another is a massive change in pace. Once I get a handle on it (though how long that will take I don’t know!) I do intend to step out again and and starting reading more similarly to how I learned Swedish. Listening to an audiobook completely without text, for example. Learning from that. Trusting that my brain will eventually do the work and figure it out.

Anyway, some more nuance for thought.

9 Likes

@nfera going back to the topic, I was trying a bit more Language Reactor that I have since a long time on YT (and that I would eventually use it as a “plugin” idea to the main method, like I was writing somewhere above).

I have seen these pros and cons, maybe you have some solution for the cons.
I haven’t looked into the Premium features.

Pros:

  • It seems it keeps me more engaged if I want to watch a video, so I remember to watch it with intention. If I just listen to the video, I import the audio on LingQ + Whisper.
  • It might be good for content I enjoy but I’m already in the confidence zone with the language so I don’t have to pause many times.

Cons

  • I cannot select more than 1 word. If I want to read the translation for a phrasal verb or combined words, it doesn’t give me the option.
  • I use the spacebar to pause the video to look up one word because it doesn’t always pause if I overlay with the cursor on the sentence.
  • I can only have the translation in one language if I click on the word and doesn’t allow me to switch languages on the go.
  • I can only translate, for example in English, a word as a verb and doesn’t give me the translation for substantive, adjective and so on if that word has different options. It doesn’t understand the meaning of the single word inside a sentence.

LingQ related:

  • Is there a way to “favorite” the word we want to, then just take those group of words and import them on LingQ with the translation and sentence reference?
  • Have you found a way to connect your work on LR by importing stuff from there to LingQ?

Thanks.

2 Likes

Hi @GMelillo,

Here are a few ideas reg. “use operations” (including L1->L2 SRSing, which is a tougher version of “active recall”):

  • Self talk (which includes recording myself).
  • Answer questions in texts (see, for instance, the end of LingQ’s Mini Stories).
  • Write summaries of what I’ve listened to / read before.
  • Write a diary about everyday events / hobbies / topics I’m interested in (say, AI).
  • Chat with generative AIs à la ChatGPT.
  • Chat with natives (see, for ex.: https://hinative.com/ or this VRChat video from Mattvs Japan: https://hinative.com/).
  • Use the L2 in games.
  • Write blog posts in the L2.
  • Take creative writing and other online courses in the L2 (e.g., on Coursera).
  • Use Italki & Co for speaking.
  • Go to Meetups in my region.
  • Practice shadowing (see, for ex., this YT video by Alexander Arguelles: https://www.youtube.com/watch?v=130bOvRpt24)
    etc.

I did most of this in the past.
Unfortunately I’m often in a hurry at the moment, so I just stick to

  • L1->L2 SRSing (including: shadowing and writing exercises).
  • Chatting with ChatGPT4 (in Bing) in various languages in my day job.
  • Self talk (for summarizing topics).

And, of course, I get free English / Spanish “sessions” in my company / family as well :slight_smile:

My main point in this context is not to separate the input (language recognition) and output (language use) phases too much (over many months or even years).
After all, trying to improve one’s language use by relying on more and more recognition operations (= Comprehensible Input “purism”) is almost as bad as trying to improve sprinting by jogging slowly for longer and longer periods of time :slight_smile:

(Note: The improvements related to output are for Dr. Krashen just a variation of input. But that’s a thread for another day).

In short, for the sake of greater efficiency, I prefer a deliberate practice cycle that more closely links the recognition and use processes.

For example as follows:

  1. Reading while listening (RWL) session in the morning (ca. 50 min).
  2. Re-listening (ca. 1-2 times) later in the evening or the next day (while doing chores, etc.).
  3. Doing L1->L2 SRSing combined with self talk (for summaries of 1)+2)), variations of the sentences on the flashcards and / or writing exercises (such as chatting with ChatGPT about interesting aspects related to the RWL session).

Hope that helps / nice Sunday
Peter

PS -
Note:
I’m not saying here that “CI purism” is not effective. I think it is (at least to a certain degree. That is: If people want to become good speakers / writers, they have to speak / write a lot. If this is true for native speakers, then it’s also the case for language learners).

See, for ex., Stevi (a fellow German) who mastered Japanese (more or less) in ca. 7000 hours.
He is one of the CI “purist” success stories in the Japanese language learning community (but he used Anki obsessively - and if I remember correctly that included L1->L2 operations as well):

I admire Stevi, I really do! But I want to achieve what he achieved (reaching an adv. Japanese level) in a few thousand hours less :slight_smile:

4 Likes

@davideroccato I don’t think Language Reactor deals well with multiple languages. Personally, I haven’t tried much.

I haven’t had any experiences where the cursor did not activate the auto-pause. Another way to do it is to press ‘a’, which is the hotkey to go back to the previous subtitle. ‘s’ repeats the same subtitle.

Yeah, that’s not possible, I think. In those cases, though, generally the translated sentence picks up the phrasal verb, collocation, etc. The the translated sentence + the option for three definitions for a word on Language Reactor means I understand the vast majority of new words in Italian. I understand much more than that on LingQ with low-frequency words only having one single-word DeepL definition. I imagine the translated sentence would be much more likely to pick up, say, the trennbare verbs in German.

@GMelillo One self-talk, which I like, but unfortunately I don’t do enpugh of, is: at the end of a lesson/book chapter, etc., talking to yourself and summarising the chapter. To appear less weird while out in public, I put my headphones in and pretend I am talking on the phone (but my headphones are really not plugged in. This way you are also trying to drill in the words you have just learnt. The topic and words are fresh in your mind, so it makes it easier. It also gives you a topic to talk about.

You can also add in your thoughts about it. Eg. “I thought this chapter of Harry Potter was more exciting than the last one. There was more action and Harry was complaining less. It started off…” etc.

5 Likes

This is fascinating! Because I was an odd kid and at age 8 my favourite thing to do was read dr who (reading age 11) with a dictionary, I just loved reading everything I could get my hands on. My brother also read well above his reading age. He listened to conversations and radio, or read complex books and estimated meanings . He used much longer words than me, despite being only 6, with hilarious consequences at times. So perhaps both work? I have always felt he was the better language learner (better able to listen and speak “live communication” in a new language), so I have set out specifically to try to read for understanding, not word meanings. I try to spot patterns of word groups, prefixes, verb conjugations etc. to work things out (I love Olly’s teach yourself stories for this). In a conversation if I focus on a word I don’t understand I lose the thread.

I am curently upper A1, so just beginning.
I like to translate and memorise pop songs (ie sonething I find fun) and listen/sing along repeatedly. Then I notice new sentence structures/ grammar patterns/word meanings I have just learnt jumping out at me when I listen to the songs or vice versa.

I truly think that finding at least some things you can bear to repeat over and over that have decent content is really helpful (my son and I are now watching the peppa pig episodes all over again -repeatedly- in Dutch).

2 Likes