The difficulty you're facing is in this:
“forget textbooks, you need comprehensible input!” which I agree with …
'Comprehensible input' is a grift. Everything I'm saying in this comment is to help you avoid the mistake I made – I listened to the grifters and wasted months stumbling my way through Mandarin and then Spanish and learned next-to-nothing. Don't fall for the grift. You already know why it's a grift:
how can you begin learning the language to the point at which this is possible?
It is highly unlikely that you will get there solely by 'comprehensible input' in the way that it is touted by internet polyglots and snake oil marketing. You need something to make the input comprehensible. Which means you need some form of scaffolding, especially at the start. And if you're using scaffolding, you are not doing 'comprehensible input' (CI). CI is not a method that autodidacts can start with (unless, maybe, you have years to dedicate and more patience than anyone I ever met).
CI was popularised by an academic called Stephen Krashen. He was rather specifically talking about face-to-face learning, where learners would have someone available to make the content comprehensible. He holds up his hand and says:
Das ist meine hand.
If you understood that message, you learned a bit of German. Good work! How do you replicate that when there's nobody to walk you through it? When there's nobody to point at a clementine and say, 'eine clementine'?
There are some methods, such as 'total physical response', used in classrooms and immersion schools that are designed to take their students through these steps. (If you pay someone enough and they know what they're doing, you can almost certainly scrap the face-to-face element and do this online, too – but where and how are you going to find that and pay for that?)
There are graded readers and a plethora of online courses aimed at reproducing the same thing. Much of this, however, is only in the written form. In fact, Krashen is also talking explicitly about 'extensive reading' as a way of getting your daily Vitamin CI. He explicitly argues that language learners should not try to speak 'until they are ready' (but gives no indication as to when someone is 'ready').
What the internet polyglots forget to mention is that you can't just pick up a book from zero and understand anything useful. Some of Krashen's subjects were immigrant children in the US learning English. They heard the language at school, they received English lessons, and they interacted with other kids and adults. Their CI came after learning the basics.
Krashen made an academic argument that has been misinterpreted and misapplied. There is some useful stuff in his work. But that work was never intended as a self-help guide for autodidacts studying languages with access to the internet in a country that doesn't speak the target language.
Please don't let this put you off. There is a way. And eventually, you can improve by using 'comprehensible input'. But, as you say, you need to learn quite a bit before you get there.
First off, you need to hear the sounds as early as possible. Otherwise, you will not understand natives and your pronunciation will almost certainly be off. You won't be able to guess Chinese tones. (I am assuming that you're not hearing impaired – this advice may not apply if you are.) So a text-only beginning is no use.
If you don't practice speaking, you won't speak; although you need to replicate a good accent, so you do have to listen first. And you need to know some words and the order to put them in, which again means some pre-study. What I'm saying is that you shouldn't wait for e.g. five years before you try to speak. Start trying early on.
If you don't understand any grammar, you won't know how to extract meaning from Chinese sentences. Neither will you be able to form meaningful Chinese sentences when you speak or write. And without sufficient vocabulary, you won't know what the sentences say even if you know the grammar. Likewise for constructing Chinese sentences.
(Now, you're lucky, because basic Chinese grammar is a breeze. It's subject-verb-object, like English. And it doesn't always matter if you switch it up a little bit. It's quite forgiving. The past tense is easier than in other languages, too – so far as I understand, you can say the same thing and just add the time frame, such as 'yesterday' or 'last year', etc. (There is also 'le' or 了, but I won't try to explain that in case I get it wrong – my Mandarin isn't great.))
I can summarise a couple of points now. You need to learn enough of the language in order that messages in Chinese make sense. This will allow you to consume the so-called holy grail 'comprehensible input' to improve. And remember, the point of all this, according to the acolytes, is to be end up 'fluent'. If that's your goal, you need to build up your skills in the four key areas:
- Listening
- Speaking
- Reading
- Writing
The idea that studying only one or two of these (reading and/or listening, depending on the interpretation of Krashen) will make you fluent will result in a lot of frustration and wasted time – if you don't just give up.
How do you learn enough grammar and vocabulary to be able to listen, read, speak, and write? There are a couple of options.
- Start with a course of some kind. (Not Duolinguo – that is at best a supplement.) A proper course. It depends on how much you can afford or pirate lol. There's Assimil, Pimsleur, Michael Thomas, Teach Yourself, etc. There's also a good course by Princeton. I compiled some other resources, here, including a link to the first Princeton book: https://lemmygrad.ml/comment/357509. There is a PDF online with links to the audio.
- Read a Chinese grammar in English. ('Grammar' by the way, refers a 'book on the rules of the language', not directly to the 'rules of the language'.) Just read through it quickly and try to understand as much of the grammar as possible. Maybe read a second one. Then go through a course.
- Vocab lists or Anki (I can't get on with Anki but some people swear by it – the science behind it is compelling, I'm just more of a pen and paper kind of learner). Either way, don't do 'themed' lists; you'll mix up the words if you learn a list of fruit, for example. [Edit: you could e.g. choose ten words/characters that you came across and had to look up. You might remember the context, that way, too, which could help.]
Now, you might say, but redtea, so-and-so says that they learned Japanese or French or whatever through CI. Unfortunately, they're almost certainly lying or unaware that they're not being truthful. If you look into their backstory, you will almost certainly find that they e.g. studied a degree in the language before they perfected it through CI. Or maybe they 'learned English through CI' but forget to mention or downplay the fact that they were taught English in school.
Once you have learned some of the grammar, there is a way to 'cheat' the CI. It's called 'Listening-Reading'. You choose a familiar book, then you read it in English while listening to it in Chinese. (Really, you end up reading the 'next' sentence in English while the narrator is pausing between the sentences – there's a knack to it, and it's a lot easier than it sounds.) Ideally, you have a parallel text. With Chinese, you would ideally have three columns: English, pinyin, characters. You can use the Princeton book for listening-reading. Or something like Streetwise Chinese. (Subtitled movies don't count.) I wrote about the technique, here: https://lemmygrad.ml/post/401145.
[Edit: bilingual texts are also very good for this kind of thing but you want to make sure that you are not just guessing the pronunciation. So maybe leave them till you've listened to enough that you know what the words should sound like. Your (internal) pronunciation doesn't have to be perfect. It just needs to be based on what you've heard.]
This isn't to say that you can't do some enjoyable things in Chinese while you're studying the grammar or course. You can and you should. If you read a chapter in a grammar or course and then watch some TV or read some news, you will spot what you learn 'in the wild' and it will go in a lot better. But the 'comprehensible input' can only come after you have already learned something.
If you're interested, Paul Nation wrote a good guide, 'What you need to know to learn a foreign language (English)': https://www.wgtn.ac.nz/lals/resources/paul-nations-resources/paul-nations-publications/publications.
Edit: so many typos! Hopefully I caught them all. I also added a couple of other points in square brackets.
It also strikes me, reading back through this, that there's a dialectic going on. Maybe the issue with CI is not CI itself, but the way that it's sold, as something that can be isolated from explicit study.
I do think it's important to engage with native content. But as part of a more rounded approach, and something to increase as you go on. A bit of grammar, a bit of vocab, look for both in the wild. A bit more grammar, a bit more vocab, look for them and the first bit of grammar and vocab in the wild.
Paul Nation, btw, argues that you can start to do CI when you understand 98% of the text; and in the meantime, to use word lists to build that initial vocabulary the fastest (I imagine Anki is good for that if you like the software).
Eventually, you know enough grammar and vocab that you'll spot things in the wild that you don't fully understand but which you recognise. I've heard these called 'known-unkowns'. When you start to see them, you can work the other way round: engage with native content and then look up the known-unkowns in a grammar or dictionary. For me, this only works efficiently when I'm filling in a few gaps, but maybe people with better pattern recognition will find more use in this approach earlier on.