Organizing and integrating notes

Can someone help me figure out where to go from here?

When I came to Obsidian, I didn’t start with a blank slate; I started with an existing (large) database of notes, and didn’t have any previous knowledge of things like atomic note-taking. The whole thing was more like a place that I can collect things and be able to find them again. (I mostly make connections in my head, so I hadn’t really had any need of making them externally; I’d make them INternally and then just need to be able to relocate the piece of information I remembered having seen.)

Anyway…so I now have this vault with over 10,000 notes. Some are quotes or web articles or other random bits and bobs like that, but the majority are either book notes (notes I’ve taken from various books I read) or they’re my own posts and comments copied from Facebook, so they’re essentially the existing start of evergreen notes.

But none of these notes were written with the aim of “one idea per note,” and some of them are overlapping (because they come from similar conversations that took place at different times or in slightly different contexts), so now I’m wanting to “clean up” some of this, and get them better integrated and organized – merge notes that are largely redundant, atomize anything that needs to be atomized, etc. – and I have no idea how to go about this.

My first thought is to start by creating notes for the higher-level ideas, and create a quick-and-dirty outline of my thinking on each topic so I can look to see which notes I have that fit into those frameworks. Is that too top-down to maintain the spirit of how atomic notes are meant to work, or a sane way to go about identifying what already exists in the vault?

Is there a better approach? Has anyone else already done this? There’s just so much here it’s a little overwhelming to think of taking it note by note and trying to figure out what might need to be merged together with it from the bottom up.

Here’s a concrete example: I created a new post today. This was my largely off-the-cuff FB post. This is what I initially have as a note:

I think we would do well to recover the biblical language of “body” and “congregation.”

“Church” becomes very ambiguous and people routinely equivocate the local church with the universal, the visible with the invisible. (And many assume that the opposite of “local” is “invisible,” which is not true.)

Within my vault, I already have a number of notes about church – what is it, how should it function, etc.? I also already have a number of notes about using biblical vocabulary – e.g. “apostle” vs. “missionary,” “preaching” vs. “teaching,” etc. So the content of this note clearly connects to both of those overall trains of thought.

The question, then, is: is this a single atomic note, which should link to both trains of thought in other notes? Or is this two ideas that ought to be broken apart because the “what language do we use?” and “why does it matter?” are more separate than what ought to belong in a single note?

(I’m not trying to overthink doing it “right”; I’m just trying to work out what best enables interlinking in the long run.)


I was in the same predicament. After avoiding it for a while, I realized that it might be worth it to make a regular practice of reviewing old non-atomic notes and, using something like Note Composer, extracting the key headings or blocks that can exist as useful independent ideas for future use, leaving behind the embed. I actually use a Templater script that adds the source text into a source heading, leaving behind the embed to that. I appreciate your feature request Note Composer: links to blocks and headers should be updated when extracting , and understand the friction and/or limitations of doing this en masse.

At the very least, by doing this, these underlying ideas wouldn’t be hiding behind a note name that cannot fully explain the contents. And using MOCs and properties, you can feel less of a need to have everything linked perfectly because you will have decent clumps. Not sure this answer is anything you haven’t already considered, but just wanted to share my experience because I was thinking the same things.

The review process becomes part of the new normal. I have yet to fully catch up and almost can’t imagine what it will be like when and if I ever do. Good luck!

1 Like

There is no prescribed way that you must organize your notes. You don’t have to use linking or atomic notes.

What is working in your vault right now? What is not working? Where is the friction? What do you want to be able to do?

I would not do a wholesale rewrite of your notes. Start linking things that you want to be linked or making up a few MOCs/indexes of subjects that you want to organize or explore.

Be purposeful about your note writing as you move forward and see what works for you. Prune out notes that you don’t need anymore. See how your content starts to organize itself as you use them.

I must not be asking this well. I’ve been just taking notes and trying to see what works. And what I’m asking about is the friction that’s emerging, that I’m having a hard time sorting out how to refine.

As I’ve thought/talked through this I think part of the problem might be that many of these notes inherently are, themselves, connections – they started out that way – between ideas that don’t necessarily exist within my vault as individual ideas. (I just “assume” them, I guess?)

So, for instance, I have half a dozen notes on the same basic topic. They’re very similar, but not identical. Naming them becomes difficult because how many different names can I come up with? Linking to them becomes difficult, because which of the half-dozen do I link to? Linking between them is difficult because there’s really not an organic way to do it; I’d have to just list all the other similar notes as “related” – and then I have alllll of these to update any time I add another one. None of that makes sense from a structural standpoint.

But I don’t know how to figure out what should be combined vs. what should be broken down or refined so the notes are different, or if I need to actually be creating notes for them to link to as connectors.

Because now I’m wondering if the problem is not so much what is there, but what isn’t. It doesn’t make sense to me to create notes of what I already know without thinking about it, because I don’t need help remembering it. But if those are the ideas I’m linking with my newer notes, they then have nothing to link through within the vault.

I struggled with this same thing in math in high school. Geometry proofs gave me fits – I could explain the whole process in my head of how we started where we started and ended up where we ended up, but I couldn’t find the right degree of granularity to know where to break it up. Where one step ended and the next one started.

1 Like

As I’m further thinking through this, I feel a little silly saying it this way, because “atomic notes,” by definition, reference atoms. But I think much of what I have going on is that I have “molecular notes” – and there are no related atomic notes, so the connection points are missing.

So it’s as if I’m writing about water (H2O) and carbon dioxide (CO2), and those feel like they’re at the simplest level. But there’s no “oxygen” (O) note, so even though I know in my head that water and carbon dioxide are related because they’re both built on oxygen, there’s not that midpoint to link them both to inside the vault.

1 Like

I would recommend placing all your existing notes in an archive folder in Obsidian. I would then create a new notes folder to which I would begin adding notes of a more atomic nature. This might be entirely new thoughts or consolidating thoughts from your old notes.

The tough part for me with longer notes is extracting the ideas contained in them. This requires some hard thinking which, to be honest is the most important part of note taking anyway.

This way you can keep what you have and build something new. If you end up consolidating notes from your archive you can then delete them and in this way slowly process your notes over time to your new notes folder.


Not entirely onboard with its importance. And most folks do not have the discipline, mindset, or need to go the extra mile; Luhmann had to, but we (because of modern software like Obsidian) don’t.

This is where the use of the word “extracting” in the quote is relevant. For me, each extraction has unique properties & meaning, and, therefore, a new context. And I do not want to taint the original within its context.

Isn’t section and block linking a interim solution for this?

For me, most of my longish notes have value for many varied reasons. There is context in each one; most were captured externally or are finished missives of my own. The value is in returning to them and discovering new insight. I’ll take context any day over structure.

Might an oxygen note be a MOC like summary to all if its manifestations?

On the other hand, I’m still in the process of applying properties to thousands of notes, but dang it, they only apply to a note. Quandry for sure until I start thinking about content tagging, indexing, and NLP and how they could be persisted in a core database in the manner our clever long range thinking developers are known for.

For now, I pepper long documents with links - a word here (i.e. [[C02]] ), a phrase there (ie. [[carbon emissions]] … and a quick search for “[[CO2” pulls them right up and dataview has incoming and outgoing link syntax built in.

And I backlink all the time. Sure, it would be nice if we could edit internal links in place, but is hovering over it really that much friction?

Whew, we’re all trying to make sense of knowledge tools that are quite new, but I’d wager that Obsidian will give us tools to move beyond Luhmann’s lingering dilemma.

1 Like

My first thought is to start by creating notes for the higher-level ideas, and create a quick-and-dirty outline of my thinking on each topic so I can look to see which notes I have that fit into those frameworks. Is that too top-down to maintain the spirit of how atomic notes are meant to work, or a sane way to go about identifying what already exists in the vault?

Have you had the chance to try this workflow since you posted?

Is there a better approach? Has anyone else already done this?

For writing a Facebook post, I think actually a single note is the best way to go and this is the way I approach notes that are being used to prepare some kind of published content.

1 Like

Have you had the chance to try this workflow since you posted?

To clarify, I’m not writing FB posts FROM Obsidian; I’m taking things I wrote on Facebook already TO Obsidian. They’re usually reasonably cohesive, but a lot of them are about overlapping things, so it’s creating a lot of bloat in my database.

Okay, so if I understood correctly, it sounds like a three-step process of searching, consolidating, and remixing. During the remixing phase, you are making decisions about classifying your notes and the scope of what you want to say. All of this sounds like a great writing process!

In terms of improvements, something sticks out to me: your first two steps could be combined into one. In Obsidian, when you open Quick Switcher (ctrl + o), and start searching for a file by filename, result suggestions will appear in the search bar. If you don’t see a result you like, you can always press shift + enter and that will create a new note with the title of whatever your search was.

My workflow makes use of a plugin called Quick Add which involves a little bit of familiarization and configuration… but can automate the retrieval, creation, linking, and storage of notes - that may be worth considering for steps one and two.

With regards to editing, I found a stack that I feel works quite well - Templater, Linter, Excalibrain.

  1. Templater ensures that all my new notes point to my daily note.
  2. Then I have Linter saves my H1 heading as an alias. After writing a note, I give it a summary as a H1 heading.
  3. Finally, I configure Excalibrain to treat links to the daily note as a “Parent”, and configured to show note aliases instead of note filenames.

Here is what that looks like for yesterday’s notes:

Basically, I’m able to see everything I wrote yesterday all at once, which lets me review what I did much more quickly, which is really useful.

This is what I do for new notes, let me know if you end up trying something like this - happy to help if I can. :grinning:


Thank you! Yes, that’s a good summary: search, consolidate, and remix.

The process of copying them all into a separate note is not strictly necessary, but I find it’s easier for me, personally, to keep track of which notes I’ve edited (remixed) and which I haven’t if I can just delete them from that “master note” as I go.

1 Like

Do your notes need to be linked?

Do they need to be atomic?

What is your goal in making a new system?

There are many functional reference systems that do not use links or atomic notes. I use descriptive note names and the Quick Switcher. If I can’t tell two notes apart or tell which I am looking for at first glance in the Quick Switcher, then I update the file names after reviewing them so that the next time I’m looking for them, I know which one I want. This may mean clarifying the description, adding new keywords or search terms, etc.

I don’t care about notes being atomic. Just that they are useful. I don’t care if they are linked or tagged, as long as I can find them using Quick Switcher or Search.

I have also atomic notes about one subject so I placed them into one folder and I created a summary query to list the notes without incoming and outgoing links:

TABLE file.folder as Path, file.cday as Create, file.mday as Modify
FROM "" 
WHERE length(file.outlinks) = 0 AND length(file.inlinks) = 0 
AND contains(file.folder, this.file.folder)
SORT file.cday DESC 

I think this helps to find the atoms which needs to process

1 Like

These are good questions. As it stands, a lot of my notes are redundant, so I’d like to find a reasonable way to reduce the repetition without losing meaningful content.


I came across some resources that might be particularly useful for you as you’re looking to organize your thoughts and strengthen your grasp on remembering key ideas.

Firstly, check out this framework which could be a valuable addition to your research: Diátaxis Framework. I recall someone from this forum recommending it, and it seems quite pertinent to your journey.

For a more in-depth understanding, there’s a presentation available here that delves into the details: Diátaxis Framework Presentation.

In applying the Diátaxis framework, I categorize content into:

  1. Tutorials: Hands-on learning.
  2. How-to Guides: Step-by-step task execution.
  3. Technical Reference: Detailed system information.
  4. Explanations: Insights into underlying concepts.

Use Diátaxis like a librarian: Tutorials for navigating your vast collection, How-To Guides for efficient note retrieval, Reference for your timeless content, and Explanations to link related theological concepts.

I’m still in the process of familiarizing myself with this framework, but I hope my current insights are useful to you.


The section on tutorials is :woman_cook:

This looks really great; thank you! This distinction between “tutorials” and “how-to guides” is confusing to me, though. I think perhaps because I think they’re not using vocabulary the way people use it in the real world.

I would consider a tutorial and a how-to guide the same thing – what they call a tutorial. It sounds like what they’re calling a “how-to guide” is what I would call a “checklist” or maybe a “quick-reference guide” – something that doesn’t actually have to tell you how to do the thing, because you already know, but hits the steps in order in an abbreviated form so you can reference the bits you might need a quick reminder of.

By way of a potentially silly example, I have two versions of my yogurt recipe on my website. One has very detailed instructions for the reader who’s never made yogurt before, doesn’t know what the look and feel should be like, etc. That would be a “tutorial” by their definition, I think. But I also have a shorter version that has just the ingredients list and a short version of the steps (heat the milk to 180, cool it back down to 110, add starter and incubate). It assumes the reader already knows what to do and just needs a quick reminder of what temperatures to watch for and which order the steps go in. I think that’s a “how-to guide” by their definition?

Does that sound like an accurate grasp of the distinction they’re making?


A tutorial is a practical activity , in which the student learns by doing something meaningful, towards some achievable goal.

So I might classify your more detailed instructions as a reference guide - but I could be wrong.

I’m still learning how to approach documentation. What I’ve been doing is starting by writing explainers - they’re the most informal, easy to write, and less technical. After I write an explainer, I will attempt a how-to-guide. There’s usually more detail than is appropriate and so that detail gets edited out. Those are the two forms that I’m focused on right now - I’ve written some math tutorials in the past - I’m going to see if I can dig those out to post them to my portfolio.