Advice on technical use, please

I’ve finished my research for a book (at least… well, we’ll call it finished). Now I have to sort a bunch of different files by year.
The main one of these, or the largest one, is 176,600 words of typed-up letters, notes, etc.
I was thinking of going down through this, copying each item into its year and month, and as I did so, using the comment feature to annotate each one with where I’d copied it to.
But is there a better way? It’s going to mean probably thousands of notes.

Can’t you just split the document into multiple sub-documents/folders?

If you ever need to see your full document again you can always compile (aggregate) it back together.

Seems like madness to try and cross-reference where you’ve filed something - effort better spent writing. If you really wanted to you could tag each smaller document with a Year Keyword (and then create a Collection with each Year keyword as a saved search). That would mean you could move the split documents around into other bits of your binder.

Or the most expensive way in terms of time would be to add a Scrivener internal link from your original document to a duplicate + split version.

(Scrivener - multiple ways to skin a cat since 1997)

I second the split recommendation. Pretty much anytime that you need to create a reference to a particular location, the best approach is going to be to create the smallest subdocument you can that contains that location. Then you can deploy Scrivener’s full array of metadata to keep track of it. Trying to find a location within a large file is always going to be much clumsier.


The only thing against the ‘split’ options is that these sections are going into different projects rather than different parts of the same project.

There are a couple of choices in that case.

You could duplicate the master document, put a copy in each project, and split from there.

You could split the master document, give each child document enough information to stand on its own, and drag into the destination projects.

You could set up a project as a data repository containing all of the materials, split into component documents, and either link to them as desired or duplicate the sub-documents and drag them into the “writing” projects.

You could use something like DevonThink Pro as the data repository, and link to it from the writing projects as needed.

But manually copying and pasting a 176,000 word document into sub-documents and then annotating each one? Maybe if you can hire a research assistant/data entry clerk to do it for you, but otherwise it sounds like a colossal and error-prone waste of time.


This is your best option I think, assuming you don’t have the time to granularise the research into a formal note-taking app (and then you still have the job of getting the notes into Scriv).

Split your reference document into ‘project’ folders - the structure gives you the metadata you need. You could, if you wanted, use a simple visual cue to tell you that you’ve copied the sub-doc (or what status it has) simply by changing the icon (e.g. default I haven’t copied it; Green notes icon - copied; Red notes icon - updated original) or something like that. I can’t find a shortcut for this, but whatever method you choose, learn a keyboard command to achieve it…

It becomes a bit more cumbersome if you need the same note in more than one project, but you could do a title convention on your doc to show you that - e.g. copy a note into two folders in your project, and prefix one M1: and one A1: (master and alternate, or whatever works for you). That way you know whenever you move / update an Mn or An doc that you need to find it’s variants.

Thanks, this good advice is helping me to think my method through.

I think I’ll continue as I started - in effect, using the 170,000-word 'master list" as the main central database and first copying that into its years, then doing the same with the various other projects.

My main problem is that I worked my way through a lot of folders in various libraries, photographing and often also typing out documents. Historians and librarians have dated some of these, but sometimes wrongly. So as I now go down through this record, seven years after having started the research, I’m going “Aha, so there you are, my beauty!” as I find something mislabelled or undated or misdated. Working through these records is itself straightening my head out about what I’m writing.

It’s a pity in a way that you can only use something like keywords about a document, rather than a paragraph, say; if I was able to note that a paragraph relating to file 44,322/30, say, belonged to 1916, and perhaps even was able to colour that paragraph in a colour coded to ‘1916’, it would be faster and easier.

But that’s exactly what splitting a document in Scriv would let you do - it simply doesn’t care about the size of the internals of any folder (ok there are some minor formatting considerations if you’re compiling out) but you can atomise your document as much as you like.

However, as Katharine said, an external note-taking app will be more fit for purpose if you really need to attach additional metadata to your research. I’ve not used DevonThink or Tinderbox which are the ones that get mentioned on here most often. I don’t think Evernote is good at this kind of volume.

ANOTHER option you have here is to stop worrying about ‘Keywords’ as being distinct from the content. You can simply add a Comment in the text (at the para / word / whatever) level. As long as your taxonomy is robust then it is searchable. Eg add a comment of “K:1977”, or “K:Albert Einstein”, and you will be able to find any references to Albert Einstein or 1977 by searching ‘all’ or ‘text’ and prefixing your search with ‘K:’

The comment will remain visually distinct from the text, and can either be compiled in or left out. [As an aside you can achieve the same goal with Bookmark Annotations, although you still can’t search on just these.]

Formal key wording would give you other options, but is more laborious.

Slightly off topic (as the advice above is better than anything I can offer), I’m curious as to why you have such a large single document that contains such diverse material? Was this a conscious decision, and if so was it effective in terms of accessing your research as you wrote? I would find it very difficult to put my finger on specific references within such a document.

Comments was my first thought, all right, and I think it’s what I’ll stick to - simply commenting each separate entry with the date into whose project it’s been copied.

How did I get such a large document; well, it’s the largest but not the only one. It’s my “Master List, General”, in which I noted the manuscript call numbers of groups of documents, then in many cases simply typed out the whole photographed document, and in others typed up a few sentences of notes.

I’d started with a series of projects, but found that because the research took in a couple of hundred people and a period from 1878 to 1932, I was in danger of doing the same piece of work twice in some cases. So it was simpler to try to limit it by keeping the main body of central research to a single long document - there are others for separate groups of letters or for separate people or statements or… well, you know.

I’m another who recommends splitting documents into subdocuments.
In the binder (with subdocuments collapsed) there is no difference.
In Scrivenings view, there is no difference.
In terms of annotations, keywords, structural reorganisation, collections, thinking about (indeed, looking at the binder and seeing) where things belong and how they relate, there is a world of difference. I honestly struggle to imagine using Scrivener without doing this and often chunk my documents to subdocuments (and sub-subdocuments, etc) right down to paragraph level (basically 1 unit of information).

[]Document 1[/][]Document 2[/][]Document 3[/]

There’s not much I can do with this other than make copious annotations/notes and search through them all, all the time.

Compared to…
[]Document 1[/][list]
[]Document 1a[/][]Document 1b[/][]Document 1c[/]
[]Document 2 [/]
[]Document 2a[/][]Document 2b[/][]Document 2c[/][list][]Document 2c(i)[/][]Document 2c(ii)[/][]Document 2c(iii)[/][]Document 2d[/][]Document 2e[/][/list:u]
[]Document 3
]Document 3a[/][]Document 3b[/][]Document 3c[/][]Document 3d[/][/:m][/list:u]

You can still do everything with Project A’ you could do with Project A, including viewing everything in identical fashion using Scrivenings mode and making copious notes. BUT now you can no do any number of other things by way of the subdocuments. e.g. collections, keywords, different combinations of documents in Scrivenings view, opening various documents in their own QuickView panels, linking to specific information (right down to paragraph level) from anywhere in the project (or other project), etc.

I just wanted to note, in case you hadn’t realized this; that you can add keywords, statuses, labels, and other metadata to multiple documents at once. Usually, by selecting them all in the binder and right-clicking on the selection to access the contextual menu. Also, if you want a reference, document note, keyword, status, etc… to be applied to the documents you split off from a larger one, apply that reference, note, etc… to the main document first. The documents you split off will inherit that metadata from the main document. Hopefully that will relieve some of your tedium.