Ah yeah, that would normally be good advice if you’re not using an extensive document preparation system on the back-end. If Scrivener is solely responsible for generating a “readable” cross-reference and table of content, then these are handy options. With the LaTeX template they are going to make a huge mess, because we are using the title prefix and suffix to insert code, rather than human-friendly things like “Chapter Twenty-Five”. The code we insert of course is what will ultimately generate human-friendly headings via LaTeX. So that is why the default template settings have that whole pane switched off.
The short answer is that there is a tool for doing so. You can select any number of links in the text, right-click on the selection, and use the “Update Links to Use Target Titles” command. In that case, Scrivener assumes you know what you want and will blow away all hyperlinked text and replace each instance with the binder title it points to.
The long answer is that for an automatic solution, there unfortunately is nothing, though for good reason. Scrivener internally and informally supports two different kinds of cross-reference. There is the sort where you would link from some descriptive text to a section or figure, and the hyperlink can be any text you want for that. The other sort of link is “named”, which happens when the visible hyperlink text precisely matches the binder item it links to. In that case, any modifications the compiler makes to the title will be made to the hyperlink text at each point in the document where it occurs—in accordance with the settings in the Document Title Links pane. This is how you would get “Chapter Twenty-Five: Red Book” in Scrivener’s ToC, when the link itself in the editor just reads “Red Book”.
So in short you don’t actually need that, again because we aren’t using Scrivener to replace what LaTeX already does. It’s going to put the “Chapter Twenty-Five” (or however the document class works) in both places for you, replacing the internal “scrivauto:32” or whatever identifies that heading.
But may see the problem with why they don’t automatically update. An outdated title meant to mention a heading by name is going to be logically ambiguous from any random string of text we might hyperlink to a section with. Scrivener can’t know that at some point in the past, “Red Book” used to be called “Working Title - Red Things”. From its point of view, that’s how we wished to descriptively refer to “Red Book”. So to be clear, the options in Document Titles aren’t going to solve that problem, either. No matter how you use Scrivener, it’s going to assume a hyperlink with a non-matching name is meant to be that way.
However, to again return to the main theme here—in general we don’t have to worry about that though. When I refer to a section in my writings, I never bother with naming it myself in the link, because LaTeX is going to do that for me. I’m going to just write normally, and let the autoref command produce the numbering and naming as I design it in the preamble (or let it use defaults).
The defaults, for Hyperref anyway, simply prints the type of section you are linking to along with its numerical designation. E.g. if this sentence here were to refer to another body of material in the work, I wouldn’t refer to it in my writing, but let the autoref tack on the clickable text for me, which will show up like so (Subsection 12.2.1). The parenthetical markings are added by Scrivener, in the Markup pane, which you can remove if you prefer to embed the reference more naturally in the language of the text.
If you don’t like that approach, then like I say, you can configure autoref to do different things, you can even use another package that is better for how you prefer to make cross-references for the reader, maybe even including the target title and page number. That’s really branching out in pure LaTeX usage though. What you need to know for Scrivener is where to modify the code that gets inserted around links (in Markup), and how sections are identified in Section Layouts.
And of course, if you need a more detailed level of control that homogenous automation doesn’t work well with (where all cross-references look the same), then you can always just type the LaTeX you need right into the editor.
And hopefully you can continue to do so! In my experience, with LaTeX you can usually do well ignoring the details for most of the composition phase. It was after all designed with a similar goal that Scrivener was, to put aside the formatting details until later. But when and if you do want to dip into customisation or figuring out how things are going to work technically, there is going to be a fair bit of tinkering and learning involved. The nice thing is that with both systems, once you find a solution you’re pretty much set, and you can usually go on writing in peace, knowing that when you compile and typeset, it’s all going to come out how you designed it.
