Of course, we do things like that, and I’m not telling you that you shouldn’t do them (I used to do that years ago when I was doing multiple languages on LingQ). But if you think in the long term, years, millions of words read, and tens of thousands of known words, you will realize that you could have spent that time practicing the language. Just consider a “percentage” of error, and you’ll be good. The rest is just collateral noise! Focus on the language, not on the “app”. Imho.
It depends on you, I wish to be consistent across languages, but I also do need to learn different names in different proper nouns depending on the language, Chinese is very infamously so. But let’s assume you wish to be fluent in Swedish, and you read “Helsingfors” you would be expected to know that Helsingfors is our name for Helsinki. But everything outside of like Finland is spelled the same way as in most other languages.
My tip is to not worry about it, add proper names if they’re important. I.e. for capitals or cities which vary significantly from languages you already know. It is perhaps inconsistent, but if you have accidentally inflated the word count with a few hundred, you will still mostly be on par with everyone else on your level. For written numbers I always add them, because they are very different between languages, the same way as “bread” and “broot” are different in English and German. So is one and “ein“.
I don’t think LingQ has ever publicly announced how it thinks you should mark words as known, as in which words you should ignore. I almost only ever ignore words from other languages, but mostly mark proper nouns and written numbers as known. I don’t think and I’m even willing to assert, that you can’t relate the know words count to actual levels in a language. This is simply because (at least in extreme cases) you could be really, really good at reading a language but not be able to speak it one bit or even have a good understanding of the spoken language.
I think when you mark words like I do, you get to fairly fluent literacy at around 40K-50K known words in most Germanic and Romance languages, but that also depends a lot on other things. If you know one or more related languages, you may get to fluent literacy much sooner than that. Heavily declined/conjugated languages, like Icelandic, may also need higher numbers. I am learning Polish, my first Slavic language, which has loads of declensions and I don’t think I’ll be fairly fluently literate until about 50K or so.
Anyway I think the known word count can lead to a “gamification” trap, where you just strive to maximize it at the cost of listening, writing and speaking. It has happened to me in the past for sure. It’s a good way to motivate you to keep going and perhaps get to a minimum of x new known words per day/week/month, but make sure to also work on listening and output for an optimal result.
I mean, saying ‘Österreich’ is not the same as saying ‘Austria’. So, I do count it as a known word. Tho it is the same saying “Martin” as saying “Martin”, despite of the accent. Same way goes with cities; Wien is not the same as Vienna.
Why is this discussion of “known words” important to some? To me it’s almost of no value.
For example: I’m learning Finnish and have 29000 known words, In English I have 12000. And I can say my English vocabulary much better then my Finnish. I can listen to 1 hour long podcast in English and understand most of it. In Finnish I can’t.
