
Language Log
Language log #000
Jun 1, 2025
What is a language log?
I have been using Phrasing to study 18+ languages simultaneously. I want to document my progress and learnings along the way.
Plus, if I can learn 10 of them to “fluency” in less than 4 years, I’ll have beat the FSI projection by a factor of 10.
Thus proving, definitively, that Phrasing is 10x more effective than other effective methods — let alone other inferior bird themed apps.
This is a blog series where I document my progress.
Why so many languages?
This is a fair question, and there are several reasons.
1-4 languages: I really want to learn Arabic, Turkish, Croatian, and Cantonese. I should make sure the app works for syllabic languages, right to left languages, agglutinative languages, and language families (i.e. Balkan Sprachtbund).
5-8 languages: I speak several languages to varying degrees — Dutch, French, Italian and Portuguese. I want to make sure that Phrasing is an effective way to make progress in any language, no matter the level.
9-12 languages: According to the FSI, learning this many languages would take over 40 years with the time I put in. If I can learn these to a B1+/B2-, then I’ll have proved 10x efficacy with Phrasing. That sounds like some sweet, sweet marketing. So let’s take some languages I have never studied that might benefit my existing language set: Spanish, Maltese, Macedonian, Mandarin.
13-16: There are a couple major language families that are missing from my studies (Greek, Japonic, Celtic, Uralic). At this point, the product knowledge of testing additional language families far outweighs the detriment from adding another few languages. Let’s add Greek, Japanese, Welsh, and Estonian.
17-18: Ok so Lithuanian and Sanskrit are both incredibly ancient languages most closely related to Proto-Indo-European. HOW COOL IS THAT. Apparently there’s a surprising amount of crossover and I really want to study these for linguistic anthropology-related reasons.
18+: At some point, I’ll probably work in some Austronesian language, a Bantu language, a Dravidian language and a Native American language. But enough is enough for now! I’ll wait until I have a good foundation in this already ludicrous list.
What makes you think this is possible?
Well, the general rule of thumb is you need about 3000 words for general B2 level. That would mean give or take 30,000 words to be deemed fluent in 10 languages. Over 4 years, that’s 20 new words per day, assuming you don’t pick up any additional vocabulary.
With Phrasing, I already know that 20 new words per day is much easier than with applications like Anki. I also know that I pick up the most common ~500 words already just learning other words with Phrasing.
So I reckon I can get to a pretty comfortable 3,000-4,000 words in most of the languages. However, just learning a bunch of words in Anki is very different from learning a language. I know, because I've tried.
This is where the magic of Phrasing comes in. Already, I'm making dramatically more progress in Croatian cases than I did studying Croatian in the traditional manner, and more than studying Russian for almost a year with Anki/Lingvist. I've used my limited extremely limited Arabic and Macedonian more than I ever have within a few hours of using it. And I haven't even gotten the chef d'ouvre: language acquisition.
Everything in Phrasing is built to unlock native level content for you ASAP, because that's where the real learning is.
Why simultaneously?
Personally I think it will be far more effective, and most of the memory science would agree with me. Spacing, Interleaving, Batching, Contextual Interface, Comprehensible Input, Encoding Variability, Category Induction, Memory Schemas — these are all theories that point to the parallelism actually helping with my recall and learning. See my article about the science of learning multiple languages.
I also don’t think there’s any “get fluent quick” trick. As with most things, your greatest asset is time. If I gave all of these languages 3 months, there is absolutely no way I would be able to do this.
Might it be more optimal to start with a few and do them partially in a series and partially in parallel? Yes, actually. In fact, this goes well beyond the very present diminishing returns present in the data. However for many of the later languages, I’m doing 1-5 reviews a day. I have no idea how beneficial it is, I don’t think it’s that detrimental — it’ll be interesting to see though. My rationale is that the benefit from just getting use to the patterns and sounds of the language will be greater than the detriment of introducing a bit of ‘noise’ into my studies.
Regardless, for now, it’s making it all the more fun, and that’s plenty.
What is fluent?
Great question. No one knows. It’s a pretty bad measurement as far as things go, but boy oh boy does it make for good marketing. However I can’t in good faith commit to this experiment without setting some more established metrics.
In this experiment, I’m just using the FSI levels, as they’re the ones telling me it will take over 40 years. Translated to CEFR, it’s around a low B2 (although I‘ll still consider a high B1 “fluent” in this experiment - language learning is not a pass/fail activity). Anything above that really just comes down to vocabulary acquisition. I’m sure in some languages, I’ll be a comfortable B2, others a high B1. I might even get to C1 in a few languages! Depends on where my curiosity and interest goes. I think my biggest limiter in this experiment is just going to be “sheer amount of vocabulary”.
Where are you starting from?
See my language log #001 for demonstrations of me speaking all my languages 2 weeks-2 months in (it varies by language)
I can speak some Dutch and French, enough to express anything I want to say and understand anything directed at me. They're good enough for Dutch and French people to compliment my speaking and occasionally not even switch to English. Those who have learned languages to a high level know that the real compliment is when people stop talking to you about your level altogether :-)
I used to be able to speak Italian and Portuguese to an extent. These were never very strong — if I was really stuck in a situation, I could use them well enough to get by, but it would not be a pleasant experience for anyone involved.
Arabic I studied for about 2 months 2 years back, Mandarin 2 months 2.5 years back, and Croatian for 3 months 5 years back. I could exchange very basic scripted greetings. A0 at their peak, maybe Croatian touched an A1.
Japanese: I’ve studied it very sporadically through my whole life, I’ve never even gotten to an A1. I know the general meaning of maybe 500 kanji, and can read and write hiragana (we’ll say katakana to).
Greek, Cantonese, Lithuanian, Sanskrit, Maltese, Turkish, Estonian, Spanish, Macedonian, Welsh — these are all brand new to me.
If you’ve studied these before, isn’t that cheating?
Aren’t you fun at parties? Look, I’ve dipped my toes in a lot of languages—I did my best to choose as many languages as possible that I haven’t studied.
But that is what these logs are for. I will try some written logs with audio samples, and try video logs where I do reviews for a little bit. I don’t think they’ll be interesting now, but can serve as a testament to my progress in the years to come.
How will you be deemed fluent?
I'm hoping at some point to be investigated by the Noble One himself, Mr Evil D.E.A. (and his team). However baring that, I'll make sure that there is ample genuine content online with me in the target language, preferable unedited, unscripted, and with a native. The internet is the harshest critic these days, so I'll leave the judgement in their hands.
I’m not going to do any sorts of official certifications or exams though. To be honest, I find them a terrible measure of fluency, and at the lower levels incredibly game-able. I don’t want to waste my time preparing for a test when I can be genuinely learning the language. Maybe that sounds like a cop out, but if Steve Kaufmann can admit it so can I:
Which 10 languages?
This part, I don’t really know yet. I would have expected my interest in Sanskrit to far outweigh Lithuanian, but I’m really digging Lithuanian. I’m just going to work on all of them according to my interests, and maybe in the last year pick the 10.
The 10 languages will obviously have to add up to over 40 years on the FSI chat, but with nearly 100 years of FSI projected study’s worth already on my plate, that should be easy to find.
Wow you must be really good at languages
I’m really not. If anything, I think I’m less naturally gifted at languages than most people. However, I love this shit. I can spend 14 hours a day studying languages and call it a vacation (ask me how I know).
Plus my interest has never been in any one language. I’ve never really been nearly as fascinated by a language as I have been by language. .
For me, the dream is having to study a bunch of different languages to make sure that they all can be learned faster, easier, and with beautiful tools. Oh no, look what happened.
How can I participate?
If you're interested in setting your own language challenge, or showing me up, then you can sign up for an account with Phrasing. We fully support over 120 languages, the app is 100% operational, and we have new features being added every week!
Get in touch to get set up and join us - to fluency and beyond!