Scrivener links

I want to create a link between a sentence in one document and a sentence in another document of the same project. I looked at the manual but it seems I can only find ways of linking text to an entire document. Then I tried to search the forum but I got so many entries that were totally irrelevant to my issue that I gave up.

I have two questions: is this possible? If yes, how?

This isn’t possible, as the manual indicates. I’ve shared a number of strategies for how this can be done (easiest of which is to just make your binder items no larger than a single topic, so that links are always relevant), but since you’ve indicated none of what you could find is relevant, perhaps a little clarification on what it the limits are and what wouldn’t work could help.

Those strategies would not work in my case simply because for this project I need to organize my binder the way it is now. And in future projects most likely I will face the same issue. I really would like to be able to link a small chunk of text (down to one sentence) in one document to another small chunk of text in another document. I thought it wasn’t currently possible, since there was no documentation for it. I guess I wanted a confirmation. Is it something that can be implemented in the future? I don’t write codes, so I have no idea whether this is a nightmarish thing to work on or not. For the kind of writing I do (scientific writing), it’d be definitely useful. If it’s doable, is it possible to move this thread (mini thread, really :smiley: ) in the wish list section, or should I make another post?

Yeah, it’s not really possible, that’s the problem. I mean, nearly everything is possible, but so is climbing Mt. Everest. :slight_smile: The problem is that the linking code is pre-packaged, much like the List code and Tables. You can work around them, embellish them a bit, but you can’t change core things about them without discarding them and starting over. This one day might be done—but probably not anywhere in the immediate future. It would take a long time to rebuild Apple’s text infrastructure, and to do so only to fix a few niggling issues with them (like linking targets and a few stray list bugs) is a bit of the proverbial sledgehammer.

That’s fine. I have to ask that question though whenever someone pops up looking for a way to link to a piece of a document, or how to add meta-data to only paragraphs of a document. Whenever I hear that question, it’s usually a red flag that the software is being used too classically, where the outline is a publishable outline like one might construct in Microsoft Word, when in fact Scrivener’s outlining system’s great strength is that it can cut into a document way deeper than the book structure has to go—precisely because the author needs more visibility of the work’s structure than the readers do—like linking to a specific location. Having a section which is a visible section to the reader, in effect cut up into 15 smaller and topically atomic pieces for you that the reader cannot ever see, is part of what makes this program awesome. Hence, I have to ask. :slight_smile:

But, if you’ve already made things as small as you need for how you think of a work in terms of “pieces”, there are other strategies—this is why I asked what you had come across so that I could suggest something you might have missed.

Now I am curious. What are these other strategies? I do make things radically small, down to one paragraph, it just doesn’t make sense to make them as small as a sentence for the way I think about it.

What I have come across is a lot of stuff that is totally irrelevant to my issue. Maybe I didn’t use the best keywords for my search, but at some point I had to give up, it felt like a waste of time. I didn’t come across any of your posts you are referring to, so if you can point me to those, that’d be great. You know what I’m trying to do, so you may have something in mind already.


Hmm well then it sounds like you’re already using the program to its potential. Well if you have things that small, then wouldn’t a link to the target document be enough? In my own extensive usage of hyperlinks within a project, I tend to keep sections around one screen-height of content tall—there are certainly exceptions, but I never find myself wishing that the link went anywhere other than the section itself. I’m wondering if perhaps you are expecting something different of the link, like that perhaps it would highlight the target sentence upon arrival or something? Because otherwise I don’t see what you would gain from linking to a sentence within a single paragraph section, as all of the content would surely already be visible. Most people asking for this are linking into a 5 or 10 screen-height length document and so thus are looking for a way for the editor to scroll to a point within the document—and that is where my advice to cut the document into topical length sections is more relevant. This kind of target is more like a web page, where you can insert an anchor somewhere in the target page and link to that anchor. The result is the browser scrolling the anchor to the very top of the window but nothing beyond that (no highlighting or anything).

Given your description, I’m not sure if the other strategies would be of any use to you either, since they are designed to make the scrolling part easier. If someone has multi-page topics that are already as atomic as they can get, then the usage of features like Text Bookmarks and inline annotations with keywords in them, or inspector comments which can also be used to auto-scroll when clicked upon, are valid ways of overcoming the limitation of being unable to scroll the view automatically upon link arrival. With a very small section, you wouldn’t need to scroll the view anywhere.

So maybe just using highlighters would be enough for you? Highlights can be stripped out when you compile (Transformations option pane) and so thus for your own usage. If you have multiple links coming in to different sentences, then maybe the inline annotation hint could still prove useful.

Here is an applied example. Note that under default settings, when you create a Scrivener Link to a target document, a backlink is automatically created, pointing back to the source document in the target document’s Reference pane. In this example, I’ve added an inline annotation and highlight on the associated sentence (though the mere presence of an inline annotation at the beginning of a sentence could be announcement enough). Both of these could be rendered invisible on output and so are purely editorial tools. Meanwhile the automatically generated backlink has had its Description field renamed to match the annotation keyword.

You could take another step of changing the “[Internal Link]” URL to a single asterisk. This is okay to do because the actual link plumbing is not something you can edit or destroy. A single asterisk in the URL field will add this Reference to the very top of the document’s header bar icon menu (to the left of the title), making it super easy to navigate backwards.

Note all of this is embellishment. By merely linking to something you get most of this out of the box without any extra work. Just linking to something creates a back link and thus increases the utility of the link network for inverse discovery. Everything else you do just enhances the clarity of the link’s purpose. If you never have more than one link arriving in a target document, a simple highlight on the important sentence is probably enough communication to yourself.

But, you can see how this stuff would be extra useful if you have 5 pages of text in the target document.

Agreed. Scrivener cannot create a paragraph from a sequence of sentences so this would only be useful in very narrow cases—dense poetry and the like.

Thanks for your thoughts.

I’d really like to link two sentences from different paragraphs/documents within a project because of the conceptual/theoretical links between them. In scientific writing this feature would be very useful. Given that this option is not possible, I guess highlighting the sentence in the target paragraph/document is the best replacement strategy available.

Yes, I can see where you are coming from better now. I think of hyperlinking in a more web-biased sense, where a selected phrase is one-way directional marker to content as opposed to a more tightly bound connector between two phrase-level pieces of text. Most people, when asking for more target specificity are even still thinking within the framework of operating within phrase->content, where the specificity is merely a zero-width marker at the target that elevates the content target automatically, rather than a some-width link that is bound to a some-width target. That’s an unusual application in my experience—and I do agree with you that it has interesting implications for both referencing and analysis; and I do not think that would be constricted to the sciences but could also have applications in the arts. Interesting side point: HTML does have this capacity. You can supply an ID to a container (which could then be arbitrarily small or large) and then hyperlink to that ID. The browser technology still treats the container as a zero-width target, using the left edge of the link boundary as the scroll position, but theoretically the semantics of the situation could be enhanced with a custom browser. RTF however does not (that is the underlying technology used for rich text in Scrivener) to my knowledge. I don’t think you can have a some-width RTF bookmark, but I could be wrong about that. I’m not an RTF expert.

That’s really a cool suggestion. I use Markdown and MMD a little (not much, all my collaborators live in an ‘office’/RTF kind of environment). I guess the MMD capability of Scrivener would allow that, right?

I agree with you, I believe this feature would be useful to non scientists too

I would like to get in here - just to be sure. I have the same problem or need but for different reasons. I’ like to write my master thesis with scrivener, therefore it is also a scientific work.
So far, I did organize my thesis as follows: in the ‘research’ bar of the binder I save all my sources I’ve read so far, sorted (but not split in different under-sections, therefore in one document) by page numbers. Then in the ‘paper’ bar I structured my thesis in the relevant chapter.
Now, here is what I want to do: I would like to write something in the e.g. introduction chapter and then make a quote or place a text fragment I will transform later into relevant text. This quote or text fragment comes from my source in the research bar and is supposed to consist of only one or two sentences. Linking the whole document with all cited pages just does not make sense. Moreover the copy of this text fragment from the research section is not enough, since I need to know from which source it comes from - therefore I thought of a link. Splitting the whole cited source into each page section…just seems so…just too much and confusing. (sometimes I cite only one sentence of one page - therefor creating a separate document…is that really necessary??)

Does the above mentioned indicate, that this is just not possible, or are there any other suggestions I could follow?

Thank you very much!

Numerous suggestions were provided earlier in this thread. Scrivener has quite a few ways to jump around within the scroll view, and they can all be harnessed to work with the linking system to facilitate jumping from one spot to another deeper in the document.

Scrivener Links are very handy to link documents considering that the link can be annotated to a text within a file. They are an exemplary feature of how Scrivener facilitates association. What if you took it to the next level and implemented a function that allows linking a text of one file with a text of another or even within the same file? This would allow for more direct association (recalling, concentration) of ideas, especially ideas that may be so briefly noted that breaking up files into final atoms that small would seem unhandy. Consider linking sentence to sentence, even word to word or multiple links.
As I’m an inveterate chaotic working on a bulky project, this is a feature I am gravely in need of.
Thank you for putting such a transparent tool out there, it turned out to be a worthy investment!

(As I see this has been suggested already. Now I’d like to know if there have been any advancements, since for me highlighting text or radically splitting documents would not suffice.)

There has not been any progress here mainly because of limitations in the linking system we have to work with. It’s not as though linking precisely to a spot in the text is a bad idea, if it could be done easily it would be.

The ability to link to a section of text in a PDF would be most useful.

I am using Scrivener for writing a PhD Thesis and often need to remember/refer to just a short piece of text/idea/note/research section and I’d like to be able to link to it.

My writing is broken up into smaller chunks but many of my resources (in Research) are not.

Scrivener for thesis writing is waaay under-utilised. This program rocks.

So, I realize it is now over ten years later, but I thought I’d chip in because this is a significant problem I’m seeking a solution to as well, and I think anyone doing research could resonate with it.

I’m writing chapters for my Ph.D., and it’s extremely common to have 40–100 pages of pasted source excerpts from my Devonthink database split up between separate subtopic-labeled notes within a research folder in Scrivener. As I import sections of each quote from this research folder, or paraphrase them into my writing document, it becomes incredibly difficult to jump back and find the precise quotation in the research folder. I literally have to go back and start doing searches which wastes a lot of time, especially if the phrase occurs more than once in my research folder.

What would be even more unwieldy and unfeasible would be to create a separate note for each one of the hundreds and hundreds of quotations (as has sometimes been suggested to enable direct-linking to individual sources). So I’ve resolved to collect them into notes by subtopic where each sub-topic could have 5–75 quotation excerpts.

Because I find myself needing to jump back to individual quotations all the time in the writing process… Is there any new way to accomplish this direct linking to a phrase within a note in the ten years since this problem was first raised?

Devonthink recently introduced a capability to copy citations with a source link. These source links are quite nifty because they do jump to the precise point of text within the file that it was copied from and visually highlights it temporarily in the case of PDFs. RTF/DOCX/TXT return the starting line, which is close enough. I mention this because it is a proof of concept for something that could potentially work/be developed in Scrivener if there isn’t already a way to do this.

Based on Devonthink’s direct link link structure, it appears that data is returned based on two variables after the file reference identifier:

  1. The position of the result based on the page number and the count of characters into the page.
  2. An immediately performed phrase search that jumps to the result if needed.
    There seems to be a hierarchy happening between these two variables as #1 seems to occur no matter what if #2 fails or isn’t performed.

In any event, I just mention that as a proof of concept that this can be done. What I’m most interested in is what Scrivener might already offer to solve this problem after ten years. Please advise.

If this post isn’t completely clear, there is a prequel post (by me) a bit upthread with, say, “the genesis” of the approach, to put it like that.

It might look complicated or complex to maneuver, but I use it all the time and it is actually quite simple.
And it works.

With that you can jump from any set point within a project to any other set point.
That is as close to a link that lands inside a document as it gets.

There is a much more modern thread on tactics for this, so I am closing this one, as most of the above dates back to when Scrivener barely even had what we’d now think of as document bookmarks, let alone a tool that allows for rapid navigation between unique pieces of text (search for “Quick Search” in that thread, which sounds quite a bit like point (2) in the list above).

What would be even more unwieldy and unfeasible would be to create a separate note for each one of the hundreds and hundreds of quotations (as has sometimes been suggested to enable direct-linking to individual sources).

I understand this can be a matter of taste, to some degree, but do know that is how Scrivener was designed to be used. It was not designed to accumulate large amounts of loosely related data into single outliner nodes like you describe—particularly with Scrivenings mode around which makes the notion of them being “split up” to be a nearly artificial concept.

But like I say, any further discussion should be carried over into the context of the more modern framework, rather than one from over a decade ago.

1 Like