You can write nonlinearly.

There are a lot of correct ways to step through contexts. Ever gone on a really good wiki walk?

Linear writing relies on the reader's memory to keep track of larger structures and relationships between concepts. This is intrinsically limiting. Whenever you have gone too long in a nonfiction work without explaining something, you have to reiterate it: think of the phrase "recall that" in academic writing. These reiterations are often like detours on a journey: a user must keep track of some original line of thought, while simultaneously performing a recall task. Names are a way to shorten this type of reiteration. One names a theorem to be able to use it in future proofs, viz. "By De Moivre's theorem…". But if you don't remember, you have to look it up. This is the essence of a reference system. A reference is just a name that works with an index – a system of relating a name for a thing to the place where you can find it.

Books often have indices at the back, relating names to pages. Readers of physical books can use a book's physical shape as an index! It's often easy to mark, or even just remember, where in the book you read something. Citations do the same work across publications; In a library, the Dewey decimal system does this. Unique reference IDs in the Zettelkasten do this too. And on computers, especially in the web, this is what links are for.

Once you have an easy, low-friction way of referencing and cross-referencing stuff, you quickly realie that there's no need for a canonical order in which to step through it. Sure, there are often better and worse places to start, but there's very rarely a unique answer. This is true even in older forms! Nobody who reads a textbook ever steps through the book in the same order; the process of reading a textbook is supposed to be very "back and forth."

And if you are trying to explain a collection of concepts with a lot of interrelations, it can be a relief to represent them in a more natural form.