Over the next few months, I’ll write a non fiction textbook and think that Tinderbox will help me organize things. I’ve already done a lot of my research, reading both books and articles and have a lot of notes in Devonthink. I’ve also a rough idea about the different chapters and their topics.
My plan is now to use Tinderbox to organize which citations and quotes from my research to use were in my text. I understand that I probably will find a workflow that suits me as I go along. But I also think that I could benefit from a little input on how I plan to go ahead:
On the top level, I have created two notes, Book and Sources. In Book, I’ll outline what I’ll finally write, and in Sources I’ll move stuff in from Devonthink as I decide to use them in my writing.
In Sources, my plan is to create notes for each book I reference, as well as each newspaper article, governmental report, podcast episode etc. This will eventually give me an overview of which sources I’m using.
For each quotation I pull from the source, I’ll create a separate note and link to it from the Source note. This way, I can easily find all quotations from a source by following the links.
Each quotation note will also have a few user attributes:
Technology (What kind of technology is discussed, Algorithms, Encryption, Surveillance etc.)
Context (What part of society is discussed, Public life, Education, Politics etc.)
Sentiment (Wether the quotation is positive, negative or neutral. I want a way to track if I’m getting the balance between them correct)
Devonthink (URL back to my original note in Devonthink)
At this stage, where I’ve just started to create notes in Tinderbox, I’ve got a few questions:
On a broad level, is sorting stuff into Sources (others’ thinking) and Book (my own writing plan) a good idea?
Will linking work the way I expect: Could I select a source and easily list all quotations pulled from it?
My idea with the user attributes is to have an easy way to get an overview of my quotes: Am I discussing some aspects in to much detail? Are not enough focus spent on education? Etc. Could I use user attributes to quantify how many notes with Tech A, Context B and Sentiment C there is?
Where are the notes for individual quotes best stored in my Tinderbox document? As sub-notes to their source, or under Book, for the chapter where I plan to use them? I’m leaning for the later, but am afraid that I would get an better overview if I was storing all quotes from a source together. On the other hand, the reason I create a note for each source and link that to the quote-notes is to get both: See the quotes in the context (chapter) where I plan to use them, and by following the links find all other quotes pulled from the same source.
I hope this all makes sense. Any input is appreciated!
You’ll probably benefit from using agents for “sorting stuff”
Attribute Browser can help with this too (see @JFallowsrecent discussion of attribute browser as a tool for his own writing – I venture he’s published more than any of us ever will)
The experts here will fill in their details with their own responses.
How to you plan to produce draft and final manuscripts? Scrivener? Something else?
Do images / tables / illustrations play a role – given that you’re writing a textbook, it seems that might be the case.
How are you managing bibliographical references and footnotes?
Yes, I was thinking that agents probably will come handy as my Tinderbox document grows.
Thanks for the link to @JFallows discussions, seems really relevant to me!
Ulysses.
Actually not. There might be some illustrations, but not so many or so detailed that I’ve to handle them in my writing process, at least not at this stage.
For the book I’m writing, there is no need for formal references. My idea is to use the links between my source-notes and quote-notes to add footnotes using Markdown in Ulysses.
Thanks! Hadn’t thought about that. Perhaps keeping the quotes as sub-notes to the source-notes and then create aliases that I put in the chapters where I use them.
But how would aliases help with finding quotes not used yet?
I’d echo @PaulWalters view on output - don’t overlook its constraints. If there are none because you’ve not investigated I’d do this now. Even if you’re totally self-publishing I’d look at well-formatted examples of final form that you can use for inspiration. What you don’t want to do it is arrive at a point where you’ve put a lot of work into a draft manuscript you can’t move to the next stage.
If you’re using lots of images and want to embed images to assist your writing, I suggest you embed only small screen-res images so you don’t choke the file with loads of hi-res image data only needed for output by something likely other than Tinderbox.
Splitting the writing (‘book’) from references (‘source’) makes sense. As has been noted, aliases can help notes have a presence in both areas plus it’s easy to link from note-to-note.
Consider using a numerical scale for this then you can more easily sum/plot sentiment.
Yes. I definitely recommend doing a few simple tests in a separate document to help you think how these will work in principle. You don’t have to plan everything in advance, but understanding the general mechanisms/techniques is IMO a big productivity boost. It saves you creating lots of data you find is just not quite right for the task you’d envisaged but never actually tested.
So, fun as it is to just dive in, a small amount of time doing some general structural experiments won’t be time wasted - you can even do this alongside early writing, it’s not mutually exclusive. Exactly what those tests are depends, of course, on what you are doing.
Probably because english isn’t my first language, but I don’t get what you mean with “output” in this context. The layout of the final book? Why would I risk finding myself stuck with my draft because of the final form?
I read the suggestions by @pat to go with either aliases or links. Are you suggesting I could use a combination?
Great advise!
That’s absolutely my intention, but I also guessed that I can use some input from current users. And I was right, already a lot of good ideas! Thanks!
Sorry, I didn’t mean to confuse. By ‘output’ I mean what you take from Tinderbox to the next stage. You might be doing early work in Tinderbox but then moving the file to Scrivener for further polishing of your writing. Or you might be exporting RTF text form Tinderbox to use in word. Or you might be planning to make styled HTML that you will put in a e-book, Or, … many different ‘outputs’ are possible.
But, each of the next (after Tinderbox) stages will have constraints. It might be you use highlighted text a lot but want to export to HTML but Tinderbox HTML export doesn’t support that, etc. In that case, you might decide early on not to use highlighting, though you might like it as you can’t really use that in the next stage of the writing process. If you don’t know what you will use the ‘publish’ the final form of your book, don’t worry, but do remember the different apps all present some form of constraint. This doesn’t mean there is a right ‘next’ tool - I suspect that when it gets to that decision you will find you have no or little choice of the output needed by the next app in the publication process.
Don’t interpret ‘constraints’ as a purely negative thing, but rather as choices you don’t get to make and which may limit the sensible choices of how you go about your writing. I believe people have created books in Tinderbox but I suspect all those probably got to the published point via another app such as Scrivener, Word (or many other apps) more suited to a traditional long-form linear ‘print’.
It’s unlikely you’d be completely stuck by your choice of draft style as Tinderbox is good at letting you change structure. It’s just the later in the process you discover things you could easily have checked earlier the more work you make for yourself.
Aliases vs links. Try these out, so you’ve a sense as to whether the advice of your fellow users makes sense for you. There is no correct way for all people/uses. I would imagine you will end up using both aliases and links though probably one more than the other (it all depends on your personal style and the sort of metadata you are trying to capture). It’s not hard to replace a link with an attribute value, or do the opposite, so it’s not a decision you can’t change once you start.
If you are using Ulysses, and you want to export notes from Tinderbox to Ulysses, then you might want to consider composing those notes with Markdown syntax. You can export those as “text” and save that exported file with a .md extension (e.g., My Notes.md) This makes import to Ulysses easier. Also you can set up an external folder in Ulysses and export your notes there from Tinderbox. E.g., I often put exported files in a folder in ~/Dropbox that’s specific to my project. I add this external folder to Ulysses desktop and iOS for syncing.
I understand Mark Anderson’s concern with knowing how your output, but I’d warn against putting the cart in front of the horse. Sure, it’s always nice at the end of the project to push a button and – voila! – get a finished and formatted file, but unless you’ve got a very small project or a very tight schedule, a little bit of extra work at the very end isn’t worth a ton of anxiety today.
Reading between the lines, my sense is that you’re working with a fairly modest set of sources and that you expect to make repeated use of many sources. It may make sense in this case to have a unique shorthand you use privately to refer to each source: Tac for Tacitus’ Annales, CAH5 for volume 5 of The Cambridge Ancient History, and so forth. Used consistently, these abbreviations or tags can be readily searched, and modest numbers are not very hard to keep in mind. You can make these references into links later. Conversely, if you decide to use links to connect sources with their use, you can convert that information into this sort of shorthand without much trouble, too,.
Ouch - I think I’ve been misread! I wasn’t talking about how you export from Tinderbox, but the fact that when you do there are inevitably some consequences and these can be time-consuming to resolve. If you used RTF when you really meant to use Markdown for styling? …in 00s or 000s of notes? That’s not a 5-minute fix. Therefore being at least aware of such constraints, even if not doing embracing them, is low hanging fruit in trying to reduce the stress of creating the final (output) document.
My point was, that whilst one may simply not know what comes next, it’s worth ascertaining if you can. If you are publishing using website X or publisher Y, it is not difficult to ask what they expect by way of input. Thus, their requirement is your possible design constraint. I noted the OP mentioned this is a non-fiction textbook, so issues of structure and style are likely to have more impact than if making a prose novel. I’ve been bitten often enough by late-stage non-trivial structural changes that I’ve now learned to looked ahead where possible. That’s the lesson I was trying to pass on.
Configuring your TBX for export - well, that’s a whole different ball-game and one I was deliberately avoiding for this particular thread.
It’s not, but it’s also not an insurmountable obstacle in the course of a big project like writing a textbook. Certainly, if you know that your publisher will want MarkDown, it’s reasonable to plan for that from the start.
Then again, by the time you finish the book, the publisher may have changed its mind, or you may have changed publishers. Things happen.
Now, when I’ve understood the gist of @mwra’s comment, I see both the benefits and possible problems with that approach. I usually do all my editing in Markdown, and will continue to in this project as well.
There will be a few books, dozens of newspaper articles, dozens of academic papers, some podcast episodes and at least a handful interviews I do myself.
