Forum Posts Feed
Organizing lots of thought snippets
< Next Topic | Back to topic list | Previous Topic >
Posted by Christian Tietze
Feb 1, 2015 at 04:43 PM
Hey,
Prion wrote:
>BJ recommends splitting everything into chunks of 500 or 1000
>words (don’t remember precisely) in order to aid the artificial
>intelligence inside Devonthink. Even if it was automated, I wonder who
>would actually do this? To be fair, I always thought it was suggested
>more like a thought experiment rather than an actual practical advice.
I concur: it’s crucial to keep the notes small-ish enough so you can handle them well. If a note becomes too complex, or almost a short essay in itself, it’s hard to link to it when you think it contains a particularly interesting thought. Because it’s internally complex, you’d have to create a link like “confer note X, section Y, where I talk about author Z” or something. If, on the other hand, you’ve got a bunch of notes on topic Y for various authors, you can link to each author’s take individually and also create a compilation. This compilation would replace the more complex note.
Apart from being more atomar and thus more link-friendly, focussed notes are usually easier to re-use in writing projects because they depend less on a context. (This is analytical: if something is part of an essay, the essay is its context, and you’d have to put some effort into its extraction when you need it without all the rest.)
This makes DEVONthink do its magic pretty well. I don’t know what effect splitting up an essay-like note into smaller notes has when it comes to suggesting relevant stuff in your database. Will the results be less surprising? Will they be boring, like a regular full-text search? I wouldn’t know :)
Anyway, this approach works great with Evernote and a Wiki and a folder full of plain text files, too. It’s easier to give things a name and to find them again.
The method of having a Zettelkasten pretty much boils down to this. You don’t need fancy equipment. You only need notes concise enough to suit the way you think so you can form and find connections.
Working with a (deeply nested) outline is a good way to see note hierarchies and relationships at a glance. I understand it’s not ideal to *keep* the notes. Outlines don’t replace a database. (I’m talking about 3000 notes in my case.) Outlines are great to give an exemplary view into your archive. Writing a book yourself will have the same effect. You can say outlines are intermediary texts: texts you *would* write if you took the time. Caring for the connections between notes manually will yield similar results: you learn a lot. Collecting notes only doesn’t, though. That’s why I think DEVONthink’s AI is great for discovery, but it won’t replace the work you have to do anyway, that is to add links between notes.
Posted by Paul Korm
Feb 1, 2015 at 09:23 PM
Specifically, Johnson wrote
>I don’t want the software to tell me that an entire book is related to my query. I want the software to tell me that these five separate paragraphs from this book are relevant
and
>Most of the entries are in a sweet spot where length is concerned: between 50 and 500 words. If I had whole eBooks in there, instead of little clips of text, the tool would be useless.
The chunk size depends on how useful DEVONthink is with its suggested matches.
Posted by Paul Korm
Feb 1, 2015 at 09:24 PM
Forgot the link
http://www.stevenberlinjohnson.com/movabletype/archives/000230.html
Posted by MadaboutDana
Feb 2, 2015 at 05:14 PM
Mac users might like to check out Outlinely in this respect.
Outlinely is a very straightforward outliner, with Workflowy-like support for rich text, the ability to add notes to outline items, and a nice search function (with search and replace, too).
But it also supports tags – not immediately obvious, when you start using it.
If you click on a tag, it generates a very nice view of all items with that tag - in a separate window. What a good idea! This makes it very easy to rapidly correlate different pieces of info.
For example: I’m currently translating a magazine using Outlinely. I’m adding a tag to each paragraph (@draft1, @draft2, @final). If I select one of these tags and click it, I can immediately see - in a separate window - which bits of text are still at draft1 stage, or which are at draft2. By selecting @final, I can list all finalised texts and easily copy them for input into the client’s DTP system. At the same time, I don’t have to disrupt or move away from what I’m currently working on, because the ‘tag tree’ appears in a separate window.
It’s one of the simplest but neatest tagging solutions I’ve seen in an outliner.