We use cookies to help make LingQ better. By visiting the site, you agree to our cookie policy.
United Kingdom
How does Comprehensible Input convert to using LingQ?
I "know" 3300 words in Greek so there is plenty I don't understand.
Should I be repeating the ministories and easier text until I know them?
Should I use translation, then read, then listen; sentence by sentence?
United States
I pay less attention to any data than to my own trial at reading a sample of a text and judging for myself whether it is the right level for me. When I was teaching, I used the 5-finger rule with my students. If you read a sample page, put a finger up every time you come to a word you can't figure out from context. If you get the bottom of the page and you have run out of fingers, the text might be too difficult. It is a very rough measurement, but it actually works pretty well.
United States
I am a follower of Krashen's Comprehensible Input method for language acquisition. I reread content only if it interests me; mostly I just try to find as much content as possible AT MY LEVEL to consume as much as I can. I don't make it into a tedious "study" activity, which rather defeats the purpose. Language acquisition is an unconscious process that happens as massive amounts of input soak into your brain, I never translate.
Czech Republic
I love Krashen, as well (especially Explorations in Language Acquisition and Use) and I translate all the time to make it more comprehensible. (Although I first try on my own, so there are different attitudes to the same mean).
And I completely agree with you about the repetition. I would be bored. I just need novelty :)
United States
What do you mean that you "translate" all the time? I have worked really hard NOT to translate but to take my target language and go straight to comprehension without my native language getting in the way.
Czech Republic
By that I mean I translate every single word I don't know. It would be extremely difficult to try to guess all words just out of context. Maybe even impossible. How would you guess colors for example? 🤷♂️
I read a sentence and if I don't understand I translate the unknown words. If I still don't understand I translate the whole sentence. It is often idiomatic for instance and there is no other way to get the meaning of that :)
United Kingdom
OP: Sherry in the above post used the three most important words in comprehensible input methods: "AT MY LEVEL".
Keep repeating/inputting stuff at your level until your level is too easy and then try something a little harder.
Personally I start with TPRS youtube channels then ratchet up in difficulty a step at a time usually in one month increments.
United States
Yes! AT MY LEVEL is the key here. I know I am finding success when I can read or listen to content and I am not even aware of language. I'm just understanding! I also listen to many podcasts outside of LingQ and am starting to listen to audiobooks from Audible and Scribd
United Kingdom
Thanks all. All the comments are welcome.
My updates current approach is:
I'll find things through https://www.lingq.com/en/learn/el/web/library/search/my_lessons
This seems to give me a variety of articles, ministories, book chapters etc.
I have macros on a mouse and keyboard to bring up the next sentence, translate it and Test To Speech it quickly.
Animal Farm and "20000 leagues under the sea" are hard for me. I'll vary content. YouTube's Easy Greek is still hard but I might try more.
I wonder if Greek texts are a little hard to come by.
I've tried importing articles, recipes, bought books.
I am motivate by my (liberal) known words count but recognised a big driver is words read. At thousands of words a day, I hope I'll sense reasonable progress.
Actually, it's a reading strategy with listening as secondary, but there since I'm hearing as I read.
Thanks for the ideas.
For LingQ, I'd like to
Australia
filtered by beginner 1, or whatever to make comprehension viable.
At your level, your search range should probably be from Beginner 1 to Intermediate 1 or Intermediate 2. Then just pick whatever seems interesting to you.
at <30% blue known words. There are problems with this but it's easily available for high voume reading. Actually, I'm basically calling this "comprehensible input" and that's quite an approximation.
Reading 5,000 words in a day at 30% new words can be a long day, depending how much repeating you do and how fast you select definitions of new lingQs (or whether you write them out yourself). 10%-20% of new words is probably more the sweet spot (but yeah, really depends on the length of the lesson), but just test out what works for you.
Australia
I "know" 3300 words in Greek so there is plenty I don't understand.
Should I be repeating the ministories and easier text until I know them?
Hell no. I currently have 2,683 known words in Italian and I haven't studied the Mini Stories in a while. I got bored of them ages ago. I removed them from my playlist after I listened to them 15x (an arbitrary number that I chose, cause I found them by then boring). I now study a mix between intermediate material (material designed for language learners) and native material (a few news articles, some YouTube videos, some Wikipedia entries). I alternate between them depending on what I feel like at the time. You want to regularly expose yourself to new words (in my case, native content), while at the same time soldifying your foundation and maintaining motivation (for me, easier intermediate material). I'm really juggling these three goals by switching up the materials I study. So, personally, you can keep repeating them, if you want, but my recommendation would be, if you're bored, move on.
Putting it another way, Animal Farm has 40% unknown words and Mini-stories have few. I can read high volume with fairly low comprehension or endlessly repeat the mini-stories (I have 20 to 30 repetitions on many). I mix a bit but would have thorught that this community would have strong guidelines about what works and what %known words %^ is right for different reading strategies.
A few things:
1) I've recommended it to the LingQ team already, but there's a better way to measure unknown words. It should be (unique new words) / (total words) * 100, not how they currently measure it, which is (unique new words) / (total UNIQUE words) * 100. With the current metric, you can't tell differentiate between a dense book or a long-winded blog post using heaps of words you already know. Because of this, you can't solely rely on % new words. There would be no "strong guidelines" because you can get such variability with the current metric.
2) I'd differentiate between % new words for courses and for individual lessons. Using the current metric, they differ. A course will always have a higher % new words than those of individual courses. For instance, for me, at the moment, the course The Adventures of Pinocchio has 56% new words, while the lesson with the highest % new words is 35%.
3) With the current metric, the % new words of lessons reduces quite a bit generally as you get better. Eg. Your very first lesson will have 100% new words, while your 1,000th lesson, no matter what material you use, you won't be able to find a course with anywhere near that.
4) It also depends on many other factors, etc. etc.
Personally, regarding intensive reading, which is all I do here on LingQ, cause they don't have the functionality for extensive reading, without absolutely stuffing up your stats, my general guideline for me, personally, at my current level, is to choose a lesson with <30% unknown words, unless I'm absolutely enamoured with the content.
As an example, I'm currently reading Steve Kaufmann's book. I started it ages ago, when my Italian wasn't as good, but currently the lesson with the lowerest % new words is a short lesson of 2:18 with 7% with the highest being an 8:22 lesson at 23% (as mentioned above, the length of the lesson influences the % new words). I know that some of these words will be proper nouns too (i.e. names), which don't count, so it'll be a little lowered than that. For me, I also look at the raw number and if it's over 100 new words in that lesson, I know it'll be a long lesson.
Probably, generally, a good rule of thumb is try to aim between 10% and 20%, but if there are lessons higher or lower, it doesn't really matter. What matters more is consistency, engagement, and motivation.
My ratio of words read to known words is about 100 which is higher than for other people I've seen. Do I just need to read more or does it indicate I'm inefficient in a way I can fix?
No. This does not matter one iota. As much as people love to say that you should be tracking known words, you don't really have control over your known words (you forget them and then remember them and this is outside of your control), you should just focus on increasing your words read and hours listened. These are two metrics, which you actually have control over.
Your ratio between known words and words read doesn't mean that much. It will differ on many things, including: how liberal/conservative you are in marking a word as known, how much you study outside of LingQ, what similar languages you already know, etc.. Why this is higher or lower can depend on many things and all those things don't indicate how efficient or inefficient you are. It doesn't matter. Don't worry about it. Just focus on reading more and listening more. Those are the only things that matter. (And keeping motivation.)
Germany
Good points, nfera!
"cause they don't have the functionality for extensive reading"
Hm, I've been using LingQ (v4) mainly for "extensive" reading (while)listening) in the last 3.5 years Therefore, LingQ (or ReadLang) seem to be suitable for extensive reading (while listening) as well :-)
"Just focus on reading more and listening more."
Yes, or to make it ultra short: "exposure time" is everything (when it comes to metrics / quantifiable goals). The rest is secondary or irrelevant.
"keeping motivation"
Nah. A "strong why" + "habits" (see: B.J. Fogg, "Tiny Habits") are all that learners need.
All the other stuff about motivation, feeling like it, etc. is just worthless chatter.
United Kingdom
My ratio of words read to known words is about 100 which is higher than for other people I've seen. Do I just need to read more or does it indicate I'm inefficient in a way I can fix?
Czech Republic
Well, I think that it depends on the number of repetitions. You won't come across many new words as frequently if you read the same text repeatedly (many times over and over). Also depens how "hard" you are on yourself - when to move words to known because, of course, this is very subjective. I, for example, move a word to a higher rank each time I know its meaning.
United Kingdom
So long as you are continuing to expose yourself to Greek in a comprehensible way, you will be learning. It is a long journey though and the main difficulty is staying motivated and engaged to keep doing it even when it seems you are making no progress. For that reason, finding a method that you find enjoyable and feel is effective is the most important thing. If that means rereading the mini-stories then you do that. If it means moving on to new content then you should do that.
United Kingdom
Putting it another way, Animal Farm has 40% unknown words and Mini-stories have few. I can read high volume with fairly low comprehension or endlessly repeat the mini-stories (I have 20 to 30 repetitions on many). I mix a bit but would have thorught that this community would have strong guidelines about what works and what %known words %^ is right for different reading strategies.
United Kingdom
Yeah that's a tough decision. If 40% of the words are unknown, it's probably a bit too difficult. You can of course slog through but it might be a bit too difficult and very hard to keep doing long term. I would say something less than 20% unknown is doable for me, judged based on the fraction of words left yellow at the end of the lesson. Anything more and I probably won't be able to get through it. Other people with more motivation and staying power than me might be able to do higher.
It sounds like you have milked the mini-stories as much as you are likely to do and you are getting tired of them. Can you find something in between them and Animal Farm with maybe 20% unknown words? I am currently reading Russian translations of the Goosebumps and Nancy Drew books. They are more manageable but still I have loads of unknown words in them.
Germany
"strong guidelines about what works and what %known words %^ is right for different reading strategies."
Unfortunately, it's impossible to give "strong" guidelines here because there are too many variables involved in second language acquisition processes:
etc.
Therefore, it's only possible to give some soft guidelines (in the sense of tendencies):
1) Metrics
It's not a good idea to focus on "words known" because LingQ counts everything as a different word (e.g., the same noun in a singular and plural form = two LingQ words or a verb with 20 conjugated forms = 20 LingQ words).
A better metric is the "number of words read" (here from the perspective of a native speaker of English):
If you want to reach a B2-C1 / C1 level in reading comprehension in Greek, you should read about 3-4 million words. It would be ca. 2.5 - 3 million words in the Romance / Germanic language family, and in non-Indo-European languages, you've to read ca. 4 million words (and above).
That said, an even better metric is the quality time you've spent with Greek (e.g., as "Pomodoro blocks", see my UWL links): If you just spend 2 Pomodoro blocks à 25 min per day (5-6 days per week) on reading while listening for ca. 1-1.5 years, then all your other metrics on LingQ will skyrocket.
This also means that the only relevant time frame in this context is the number of minutes / hours, but not days, weeks, months, years, or decades.
2) Reading material
a) It's a good idea for beginners (A1 / A2) and low intermediates (B1) to avoid fiction for language learning because fictional texts are notoriously more difficult than non-fictional texts (at least for a general audience).
Notable exception: graded readers.
b) At the beginning, it's also better to read shorter texts (Wikipedia articles, the transcripts from beginner podcasts, articles from simplified news for beginners, etc.).
However, the beauty of Audio Reader software à la LingQ is that it allows you to tackle more difficult texts that are way beyond your current level. Unfortunately, reading long, esp. fictional texts with LingQ can still be a "struggle" on a B1 / B1-B2 level...
So, if I were you, I'd focus on different kinds of texts right now:
etc.
Then at a B1 level (ca. 12-5k LingQ words known), it makes sense to first resort to longer non-fictional texts for a general audience such as Yuval Noah Harari's books ("Homo Sapiens", "Homo Deus", etc.) before reading contemporary fiction that interests you.
3) Reading while listening is a more effective and efficient strategy than reading alone, esp. for an extended period of time (see my UWL links)
Therefore, it's another good strategy to hunt for texts in your L2 that have an audio version.
4) Repetition
5) If you want to test your "word knowledge" explicitly, then you can export the words read on LingQ (with some context info) to Anki.
Works like a charm and I've been doing this for years.
Apart from that, it would also have been a good idea to learn the most frequent expressions in Greek with the free courses on Memrise before using LingQ's Mini Stories: https://app.memrise.com/courses/english/greek/
Your whole experience would probably have been more pleasant and efficient.
However, if I'm not really mistaken here, you could still benefit from flashcarding with Memrise at your current language level in Greek.
And as soon as you reach a B1 level, "ultrareading while listening" (where the accelerated audio speed is the pace maker for reading) is probably the most effective and efficient reading strategy there is.
BTW, if you want to reach "(oral) fluency first", then it's probably smarter to adopt a variation of Will Hart's approach (see my links), using Netflix series, Migaku / Anki + Italki & Co before using LingQ / UWL..
However, if you can live with reaching (oral) fluency later, then using LingQ / UWL first is a good strategy.
Good luck
Peter
United States
My Japanese listening and reading is nearly 100% fiction.
The only reason it's not 100% is because some of the stuff I listen to is autobiographical, which is technically non-fiction, but still stories.
I don't do it because I'm hardcore or a language mensch; I only do it because non-fiction in Japanese is very dry to me and I enjoy fiction a lot.
Most Japanese non-fiction requires wading through dates, dollar or yen figures, percentages, tons of proper nouns, family names, company names, prefectures, all to hear one small soundbite that a bridge failed during a typhoon or Nintendo exported 14% more units last quarter. All thrilling stuff!
I think Iisaku's Thinking in Japanese podcast this week covered why reading/listening to novels / fiction is superior to non-fiction.
If I remember correctly, two major points were that non-fiction is subject to skimming for information as opposed to fiction, where the reader is more immersed because they're savoring the story in its entirety as opposed to only listening for information they want to glean.
The other big point was the opportunity to acquire words faster. In non-fiction, the sentences are presented dryly and matter-of-factly. With fiction, we have the material delivered by characters who have their own perspectives and personalities. Hence, less "hits" and angles of the words or phrases are required for them to stick because we're getting a better sense of meaning from the superior context which is charged with aliveness and emotion.
Germany
"covered why reading/listening to novels / fiction is superior to non-fiction"
When we talk about the distinction "fiction / non-fiction" in SLA, we usually don't talk about advanced science. Thus, "non-fiction" refers rather to "popular science" books such as Yuval Noah Harari's trilogy ("Home Sapiens", etc.).
And the popular science literature is teeming with stories, metaphors, etc.
It's the same with travelogues, newspaper articles, etc.
So "fiction/non-fiction" should be translated into "fictional - non fictional stories / storytelling".
However, it's the quality of the stories (or storytelling) that is decisive - not their fictional or non-fictional character!
In other words, if the fictional stories are highly conventional, full of stereotypes, etc., then it doesn't matter if you have POVs or not. In short, these stories are simply "bad".
My favorite example in this SLA context: graded readers with artificial stories created with educational intent. They are sometimes not simply bad, but awful.
So the path to learning hell is not only paved with good, but also educational intentions :-)
"we're getting a better sense of meaning from the superior context"
Our minds are basically "narrative devices" and our experience relies mainly on narratives. Therefore, we remember interesting stories more easily, but it doesn't matter if they are fictional or not.
The only advantage of fictional texts is that their vocabulary tends to be richer and more nuanced compared to their non-fictional (popular science, etc.) counterparts.
And that's exactly the reason why it's a good language learning strategy to resort to non-fictional texts first and to fictional texts later (i.e., from a B2 level onwards).
Nonetheless, feel free to choose whatever your heart desires - as long as you stick to it :-)
United States
We all have to stick together in listening to Fiction & Novels. There's no team in teeming.
Germany
"We all have to stick together in listening to Fiction & Novels. There's no team in teeming."
The Amazon algorithm and thousands of Amazon customers say hello to Houston and its "te(a)eming" problems :-)
https://www.amazon.com/storytelling/s?k=storytelling
And we haven't even touched the topic of re-entries of the distinction "fiction / non-fiction, that is, fictional elements in non-fiction and non-fictional elements in fiction.
"We all have to..."
No "we"
No "have to"
No "stick together".
Welcome to the 21st century! :-)
United States
re: repeating mini stories. In the beginning stages, I personally have found repetition to be helpful to a degree (sample size of 1 language). However, I've always moved on after at most a handful of repeats of these lessons. I find I'm simply able to memorize the story and then I'm not sure how helpful it is. Plus, it's boring.
So to a degree it depends on your tolerance for repetition. There are no hard fast rules in the community. Some people do no repetition at all and have learned languages just fine. Some like a bit of repetition.
One thing I'll suggest is don't keep repeating a lesson in the hopes of getting every last word to known. There will be words that will simply be very very difficult to move to known. Do not let these prevent you from moving on to new lessons with new words. Sometimes these words don't stick. There will be plenty of words that will stick...move on and learn these words. The more difficult (for you) words you will see in other contexts and eventually they will stick. I STILL have some words that I've met early on that give me pause for whatever reason.
Also don't be afraid to be liberal in marking a word known. My preference is to mark it known once I know it in the specific context I'm reading. I've gotten very liberal about this as I've learned the language. I'm also equally liberal it setting it right back to yellow if I meet the word again and don't remember it.
Animal Farm maybe is a bit tricky at 40% unknown, but if you the story already then you can certainly factor that in as well and tackle something more difficult. LingQ does allow you to read things that are way beyond your level because of the quick and easy translation. Particularly useful is sentence mode so you can see the full sentence translation which is helpful in comprehension and with staying engaged in the story (especially if difficult).
Maybe look for something a little easier. Graded readers as Peter suggests may be good. Not sure if they have them for Greek or not? Or perhaps look in "young adult" books. Still difficult, but probably simpler. For reference, I think I started Harry Potter using Lingq when I was around 10-11,000 words known. It was a struggle at first. Towards the end it got a lot better.
Maybe there is an "easy greek news" site. Just a quick glance I saw something called "weegreek":
https://www.listennotes.com/podcasts/weegreek-short-stories-and-news-in-easy-and-0jZfq-aWTlS/
Maybe something to check out for something interesting.
United Kingdom
Thanks. I'm pretty liberal about marking words as known.
Harry Potter is about 38% unknown to me. I dip in and out.
I'm probably faced with mixing 10% and 30%+ texts.
Germany
Hi bbbblinq,
some strategies you can use can be found in the comments from Eric and me (esp. my discussion with Wylander from March / April 2022) in this thread:
https://www.lingq.com/de/community/forum/support-feedback-forum/help-me-make-sense-of-the-cont
Reg. your Facebook group for Greek:
I provided a kind of "comprehensive framework for language learning, esp. for Indo-European languages" in this thread above.
You can also apply this framework with its mix of approaches and tools for learning Greek if you include learning the Greek writing system.
Hope that helps
Peter
United Kingdom
That's a very full guide. I'm looking for specific ways to interpret Complrehensible Input and how to use LingQ.
United Kingdom
This is useful - thanks. https://www.lingq.com/de/learn/pt/web/community/post/4385442
If I use the text +audio function, my words read dont' go up; only if I go sentence by sentence. Did I miss something?
Germany
"If I use the text + audio function, my words read dont' go up"
Unfortunately, I haven't much experience with LingQ version 5
because I'm using ReadLang at the moment.
However, your words read count should also go up when using text + audio.
If it's a bug then that's a question for @zoran / the developers.
United States
Are you talking about the mode that follows along in the text with the audio? Yeah I think it's a bug (well imo it is). You can go into the stats for the lesson thought and adjust the number of times you've read the lesson to compensate.
United Kingdom
Yes.
