Certainly I have spent more time reading about how to learn languages than actually learning them — though that is changing.
My daily practice is spending 30-60 minutes every morning shadowing Chinese in the Get Speaking Mandarin (GSM) course and about 15-20 minutes reading Latin before bed using Lingva Latina. Recently I have cut all review for Chinese in order to march forward to finish the course. Realistically, I cannot get through GSM3 before the end of the year, but I will finish GSM2 next month.
The hours of actual language study are stacking up in such great numbers. Now I need to spend more procrastinating time on planning out how to study. So, I have been considering how to merge the 10,000 sentence method (see AJATT / Antimoon) with Outlier Linguistics.
I like the sheer muscle memory practice of Outlier’s GSM — their approach is just an absolute crap ton of shadowing along with sentences to develop automaticity. The unfortunate aspect is you have some set mp3 structure of a sentence repeated 20x or 100x for example — it also doesn’t play nice with spaced repetition. I suppose you can have a card with multiple audio files, but in reality why do you really need multiples when it’s just the same audio repeated a different number of times. I suppose I should ask Claude if Anki has a plugin for that, err… let me do that… it could be done with JavaScript. But I still feel limited w/ audio length.
I made my Tortoise Mode app for dealing with longer audio files, which is something that Anki is not so great for. It’s spaced repetition, but with no rating scale on it. Based on the idea that you can revisit a podcast or longer piece of media later on more than once, so you can go deeper the 2nd and 3rd time listening to content after some kind of break in between.
I have wanted to go back into studying Japanese again, and want to try modifying one of the large sentence decks, but with heavy mouth deadlifts of saying the sentences a bajillion times. A second, different app? Going to start planning.
AI Disclosure: This post was not written using AI. AI only used for minor spelling, and grammar correction. Unfortunately I am to blame for emdashes