I wanted to try the "input first" method. I don't think its actually working for me.

I started studying Turkish (TL) 4 months ago. I have studied other languages before to high levels, but always with a teacher in a classroom. I have been reading about the idea of not doing output until you have been doing hundreds of hours of input, which promises a more native-like speech at the end.

In the beginning, I thought I wasn't interested in talking to people anyway, I'm not that sociable, I just want to read books and watch videos, so no harm in delaying output.

After about 300 hours of studying, I'm not convinced that there are no disadvantages even in this case.

First, after about 200 hours, I realized that I actually started to narrate my thoughts in (very rudimentary and grammatically incorrect) Turkish. In other words, output happened, whether I wanted it or not, but there was absolutely no feedback in any form. In real life, even if the native speaker doesn't correct you, they keep responding in the correct language.

Second, doing output actually helps you remember words and grammar points. Originally, it felt like output would just replace time spent with input, but actually, I realized that I need to do a LOT more input to make up for not doing any output.

And finally, I think at some point it just becomes hard to resist doing output. At least to me, it started to feel like I go to a party every day, only to sit in the corner and go out of my way to avoid talking to people. Even though I'm not that social, I feel like we all have the need to connect to people, and sticking to rules like "no output for 1000 hours" felt needlessly constricting. And I'm surprised how motivating it can be when I say something to someone in their native language, and they respond, and I understand the response! Even starting out not wanting to talk to anybody, I realized that I actually do, even if it's just responding to a comment about something stupid.

On the other hand, I'm not that convinced of the benefits of not doing output. It promises better grammar and accent, but if that's the case, why don't people who really need native-like speech, such as those working for intelligence, delay output?

And as for accent, I actually read about a study that really resonated with me. This study said that people with the most native-like accent are the ones who can really see themselves as part of a new culture on some level, who are the most assimilated into their adoptive cultures.

Anecdotal, but it seems to be true for people I know, including myself. My English is my strongest language by far, but its also the one with the worst accent, in spite of living in the US as an adult for 7 years. I think the reason is that I spent over a decade learning this language as an uninteresting school subject that is pretty much removed from an actual culture. It was like "I have to learn how to use present perfect for the test tomorrow, and then I can go and do stuff I actually care about," instead of a genuine interest in a culture. By the time I actually started to regularly interact with native speakers, instead of my Hungarian-accented teachers and classmates, I had been exposed to mostly the wrong accent, and practised the wrong accent for over a decade.

If this is true, interacting with native speakers early on is actually going to benefit me more than hinder my ability to speak like a native. But at this point, I'm thinking that even chatting with AI is better than completely delaying any output, because output has so many benefits, even if you truly don't want to talk to anybody.