One of the many reasons why doing a PhD is so hard is that there is no single, consistent method of teaching PhD students how to research and write.
If you want to learn a language, for example, there are systems in place to help you do so. People have figured out how to break languages down and help you practice (and, crucially, those teaching systems are shared). You don't have to figure it all out yourself (although you can), and teachers don't have to invent their own way of teaching because there are frameworks they can use. The same is true for learning to drive a car or perform brain surgery or countless other skills.
For PhD students, the guidance you get depends enormously on your supervisor. But they've most likely had very little training in how to manage and mentor you. The way they do things will be determined by their personality and the experience they went through.
There are some superb supervisors out there. It's just a shame it's such a lottery.