I wasn’t sure how best to phrase this request, but when I’m building a complex structure of chapters and the text segments within each chapter, it seems like I need to do the same to keep track of research. I would much prefer a way of viewing research that is relevant to the specific text segment I’m working on, or perhaps the research for the chapter I’m working on. I tried searching for this in the forums and found some different ways of referring to this, such as associating research with documents, but I didn’t find too much about this idea.
A sort of ideal interface for me would be if there was a small drop-down menu next to the title ‘Research’ that let me choose either ‘All’ or some number of levels to show. For example, if I’m working on a specific text segment and have the level set to 1, it would only show research associated with that document. If it’s 2, then it would show the research for the folder above that document and everything in it.
An example:
Draft
Chapter 1
[list][*] Text 1
Text 2
Text 3
Folder A
[list][*] Text 4
Text 5
Text 6
[/:m][/list:u][/:m]
[*] Chapter 2
Text 7
[/*:m][/list:u]
In this example, if I select Text 5 and have the level set to one I’d see the research for just Text 5. If I set the level to 2, I would see the research for Folder A and everything in it (Text 4, Text 5, and Text 6). If I set the level to 3, I’d see the research for Chapter 1 and everything in it. If I set the level to 4 I would see everything.
The great thing about this is that it wouldn’t need to change anything with how people work now. People could continue to work the way they have and leave the setting at All. They could also choose to associate research with specific documents (even multiple documents) and then use the levels as a way of keeping focused when working on a specific document. Associating research with a document could be as simple as dragging it from the Research area onto a document.
One could argue on how to best define the levels. Perhaps level 0 should be the document you have selected and level 1 should be the level above it. Maybe there can be a ‘current’ level and +1, +2, +3, etc. This can be hashed out a bit more through discussion (or just decided on by Keith).
Would outline hoisting solve the scope issue sufficiently for you? Hoisting will be arriving in 2.0, and it works very simply in that it will take the selected container and blot out everything else in the Binder, showing only that fork and all of its descendants. So in your example, if you selected Folder A in the Binder, and invoked the Hoist command, the Binder would become titled “Folder A” and all you would see in the list are the three text files beneath it.
Another thing that might help you out, which will be coming in the next version: You’ve always been able to create a cross-reference from one item to another in the References pane, but it’s been made much easier to access this item if you set it up right. Here is an example:
Say you want to associate a particular folder in your Research area with “Text 3”. You drag that folder into “Text 3” Reference pane, and then change the “[Internal Link]” designation to “*”.
[size=9]Internal link changed to asterisk[/size]
This will add the item to the header bar icon menu for “Text 3”. Now you can just click on that icon and the research folder will be there, letting you quickly load it in the other split as a Corkboard or Outliner.
[size=9]After clicking the ‘Characters’ reference[/size]
Multiple items can be designated with an asterisk, so it would be easy to flip between them while working—you wouldn’t even need the Inspector to be open.
I know none of these quite match what you are suggesting, but it seems to me they might approach the problem in a tangential way that accomplishes the same concern.
I’m not sure these really address the issue of associating research with documents and folders. I know you can add research links to the document notes, but that’s just for the document, and not for the level above it. Thus if I want to see all the research for a whole chapter, I’m out of luck, right? Also, the document notes do not allow the same kind of hierarchical organization that the research section does.
One thing which is weird also is that I noticed if I drag a reference into the document notes, it adds a link to that reference, but if I click on that link it loads the reference, but doesn’t select the reference in the References section. This might be by design, but I find it makes it hard to know exactly what I’m looking at…
Does outline hoisting have anything to do with associating research with documents? You’re saying it will make it easier to focus on the research you’re looking at, but it doesn’t do anything automatically, does it? I still need to build parallel hierarchies for my documents and research. Does hosting work in both the documents and research sections?
What I really want is to not have to organize my research in parallel to organizing my documents. It is a serious duplication of work. My methods eliminates this duplication and keeps the research organized without too much effort.
I must admit I’m having a hard time visualising exactly what it is you are wanting to accomplish. Going back to your original phrasing, what is it that you mean by “…it would only show research associated with that document”. I took that to mean you were dragging references into the reference pane (not the notes pane), as this is really the only formalised method for association in Scrivener. Links can perform a sort of association, but they are much more free-form, and more like the way two web pages interface with each other. You can establish a complex network with Scrivener Links, but as with the Web, there is no governing system of “association”, unless you count external parties like Google. They have a disadvantage (for this purpose) in that associations cannot be quickly seen in a single glance, since they are scattered about contextually throughout the text, whereas references sport the entire thing in a concise list. Where text links do shine, however, is in situations where having an overall list isn’t important, and its the context within the text that leads to a desire to branch out to other documents (again, like web pages).
It sounds like you are establishing associations via Scrivener links (created via drag-and-drop into text fields) in the notes pane. This is fine and dandy, and quite useful as it allows you to fully annotate the links, but references are probably more useful for establishing association.
Continuing on, it sounds like you want the Binder to selectively hide and reveal elements based on what you click on within the Binder. In my mind, that would lead to a world of confusion, where you end up in “tunnels” you can’t get out of without reseting the display mode. It would require meticulous configuration from each user, to work properly. It seems to me a more fluid system utilising existing non-invasive tools would be more generally useful—but I might be completely misunderstanding what you are getting at.
This is where I get confused, because research is documents, of some sort. Hoisting isn’t automatic in any way though. You select a container and blot out the rest of the Binder. It’s analogous to the hoist command found in dozens of other outliner based programs. Hoisting does work all over the Binder though, it isn’t something you can only do in the Draft. However, again as with most hoist tools, it’s single-container based. You couldn’t hoist two containers at once.
I think you are right though in that hoisting isn’t quite what you are looking for, since it isn’t one part of the binder you wish to focus on, but two or more parts.
The future feature, Collections, might help you out a bit here. Collections let you dump documents from all over the Binder into a non-linear arrangement so you could put all of Chapter 3’s documents in there, and all of Chapter 3’s research and see it all on a single corkboard. You could even do this automatically with keywords. If all of Chapter 3 and all of Chapter 3’s research documents have a common keyword, you could perform a search for that and save it is a dynamic collection. The only downside there is you would lose the ability to spatially re-arrange the items since it is dynamically generated. That may not be important to you though.
You originally posted only an example Draft, are you saying that your Research folder looks exactly like the Draft folder, only it has research material in the document items instead of draft text? This seems limiting to me, what if you need a dozen research articles associated with a single “Text 3” document in the draft? That’s where References come in, as they allow a single-point nexus to non-linear organisations in the Research section. Research can be organised topically, and linked to dynamically and fluidly with references. Text 6 and Text 3 can both utilise the same reference file, if need be, without hierarchal duplication.
Apologies if I’m still off-target. I think I only barely understand what you are getting at. I think I see what you mean about having two parts of the Binder with identical hierarchies, but I don’t quite follow how these would be automatically associated? What if you had three or five identical hierarchies, would they all be automatically appearing and disappearing as you click on stuff?
That still doesn’t solve what you seem to be expressing a desire to do: somehow have hierarchy without making hierarchy. Your primary complaint is that you have to build an outline twice, so how does any kind of interface binding between two hierarchies solve that? Where would this research material go, if you aren’t organising it somewhere?
Not that it will change the feature request you are advancing here, but I wonder if you have used devonthink. I does what you are talking about automatically with the see also and classify features.
I use the research folders in scrivener for media and compiled research that needs verbatim quotes or descriptions. Devonthink, on the other hand is where I get the contextual hooks into the research library I’ve built around the topic.
As I’ve been going along in my project, I’ve been taking a section I wrote in scrivener, copying it to devonthink an using devonthink to analyze it and bring my relevant research to the foreground for me.
It’s the best use of these two very powerful programs for my habits and the same desire you express in not duplicating work.
What’s been nice is to see Devonthink recognize what I wrote an have the same research pieces come forth as most relevant. What’s been nicer has been when devonthink looked like it was getting it wrong, but as I read the related materials it suggested, realized that I was overlooking something.
Durn nice to have a law clerk in my off that doesn’t smoke and work the same hours as I do. Now if I could only get one of those programs to answer my email intelligently for me. As it is, they dot say much.
I’m not sure if that quite solves the problem either, in fact it might even be worse because what the original poster wants is a completely automatic solution within Scrivener where if you click on one Binder item, the whole rest of the Binder just disappears according to some level setting, revealing only research associated with that item, or selectively, the items above it. Having a separate outline in another application entirely would be more work than having an outline in the same application. I don’t think the AI stuff would be precise enough here. They want a very specific set of documents shown dependent upon what is being clicked upon. Otherwise, a semi-fuzzy network of References would suffice, I think.
Let me re-state what I’m looking for a bit simpler. Imagine you’re writing a long book that entails a lot of research. Let’s say you have ten chapters in the book. Each chapter has a folder in the Draft binder. In each chapter let’s say you have ten topics that you discuss. Each topic is a document in a chapter folder. Overall you thus would be researching 100 topics in the book. No matter how you organize your research in the research panel, a hundred topics is a lot to keep track of, and you might have dozens of links and articles for each topic.
What I’d like is that when I’m working on a specific document, that I can limit the view of research to the research for that topic specifically. If I want I can then expand the view to all the research for that chapter (which is one level up).
Like I said, this wouldn’t require any change in the way people currently work, it would just offer an additional option. If people choose to link research to specific documents in the Draft binder, then they can utilize this feature to limit the view of research to that document, or however many levels they want. For complex works I can imagine sections of the books, with chapters, and sub-sections in each chapter, and organizing all of that research and viewing what you need is difficult, but this would make it easy.
I’m wondering then if this is something that could be done with the collections feature in 2.0, where you could use keywords to keep track of the research topic as well as the chapter. Naturally you’d have to assign the keywords manually, so if you moved topic A from chapter 1 to chapter 2, you’d need to grab all the “topic A” research and replace the “chapter 1” keyword to “chapter 2,” but this would give you a way to call up the research for a specific topic as well as all research for a specific chapter. Right now there isn’t a hierarchical way to do this with the “level up” sort of outliner approach you’re suggesting, but you could use saved searches to get this now and it looks like the collections feature will be a classier way to do this in 2.0. (For that you may not even need the keywords, but I suggest them in any case since it allows you another method of organization and also a quick way to ensure you’ve kept your collections/saved searches up to date–provided you tag your research documents when you add them or if you move the topic to another chapter.)
I realize this isn’t exactly what you want, but maybe it could help?
Have you considered interleaving your research documents with your draft documents? It requires that you explicitly exclude the research documents from being compiled, but you’d be able to block out the rest of the book’s chapters using the hoisting feature in 2.0.
Example of my original thought of interleaving research and sections of your writing:
chapter X
research1
section1
research2
section2
research3
section3
Example of putting research and your chapter sections in the same folder (better for making use of the Edit Scrivenings view):
chapter Y
research1
research2
research3
section1
section2
section3
