@zzcguns thanks for the reply, super helpful, and sorry my slow reply!
Agreed! We’re working to improve this. Lists for Indonesian and Italian have been added, we’ll work on more soon. We paused on adding podcasts and YouTube videos as those require more time and effort - someone has to go in and upload all that. We’ll see if we can’t improve the existing lists like you mentioned, getting the corresponding YouTube videos and podcasts embedded and figure out the RSS feeds, then see if they get any further tractions / it’s worth getting more videos/podcasts created and uploaded. Any additional thoughts or anyone else that finds these particularly useful please let us know.
Edit: Italian and Indonesian are still generating, they should be available by the end of today.
Looks like ~1.14 sentences per cloze. There end up being repeats as a result of translating from English into each target language.
Based off an English frequency list - for the new Fast Track, we created sets of sentences in English, then have them translated into each language. More on the rationale above.
No CEFR correlation. We are working on creating CEFR collections.
Romanian is out! Arabic - we got bogged down with trying to get harakat added, perhaps we should just release it without harakat for now and get it added later when possible.
For everyone @here in this thread - what do you think if, for example, we just release the first 3 levels of the new Fast Track, so 3 collections of 1,000 sentences each, for some languages?
Our thinking is that we could more quickly release a subset for some less popular languages, for example Catalan, Latvian, etc., that we might not get to as soon otherwise in aiming for the full set of 10,000 sentences. Then we can add more of the collections later if it turns out people are tearing through the initial subset.
We can’t think of a good reason for why not. It’d be awesome to get more collections added for more languages, even if initially not the full on 10,000 we have in mind for the new Fast Track. But curious to hear if you have any thoughts/feedback.
So if I understand it correctly, finishing all 10 collections will get me, vocabulary wise, based on the clozes only, to C1. Am I right?
CEFR collections will be great.
Since Clozemaster helps with both active vocabulary (clozes) and passive vocabulary (rest of the sentence), may I suggest the following table as a guideline for the sentences for each collection?
The numbers may vary, this is just from one source, but the rationale is that sentences with a passive vocabulary (non-cloze words) that’s limited based on the CEFR level will be more beginner friendly, make the retention of the vocabulary easier at the beginning and also make the start less overwhelming.
What do you think?
About Romanian - I’m still not seeing the new fast track in my account (I tried both on desktop and mobile).
Does it take time to update?
About Arabic and harakat - it should make it more beginner friendly, I think it will add a lot of value.
Is there any chance that in addition to the translation you can add the original text in Latin alphabet (e.g. marhaba)? It will help a lot with learning the alphabet and the reading skill and will make going through the sentences and learning the words much easier for beginners.
About your first 3 collections suggestion - I think it’s a great idea and you will be able to decide where to invest in the additional collections based in the popularity of the first 3 in each language.
What’s the source for that chart? Clozemaster counts different forms of a word as different words. So if C1 passive vocabulary counts “bebo” and “bebes” as different words in the 10,000, then it would appear so. If it’s 10,000 lemmas, then it may not, though the new Fast Track likely doesn’t have all 600 A1 words as the missing word for example, so it may be more biased towards higher CEFR level missing words.
That table is helpful for creating CEFR collections, thanks.
It should now be available, hit reply too soon Please let me know if you’re still not seeing it.
I’m rooting for Catalan to come soon - an initial 3,000 would be great!
I’m enjoying the New Fast Track. I joined up around New Year’s. I’m a native English speaker, fluent Spanish speaker. My main language that I’m learning is Lithuanian, and I’ve gotten through the first 1,000 words. They seem pretty high quality from what I know. The explanations I see using the Chat GPT feature (which I love) match what I’ve learned in grammar books. It even pointed out a few errors that I found myself (and did report).
As a point of comparison, I’ve simultaneously been plowing through Spanish - to kind of see how the collection progresses, and to identify maybe a few words here and there that I’ve forgotten. I’m through 9, 780 - so almost there. And the collection to me seems very high quality with few to no errors. (I do have a Master’s degree in the language, so while I’m not infallible I’m also not talking out of complete ignorance.)
Anyway thanks Mike, for all the great work on this. It’s a very impressive tool - seems to me it is the best tool currently available for Lithuanian in particular.
I think this is a great idea. I’ve recently started learning Tagalog and being able to make a start on the first 3000 would be really helpful. I’m not certain whether that particular language would be included in this proposal as it sounds like it’s coming soonish anyway, but I can see it being useful for others learning other languages too.
Arabic - we got bogged down with trying to get harakat added, perhaps we should just release it without harakat for now and get it added later when possible.
Please let me know if you’re still not seeing it.
