How to manage not-concrete topics

[mh_amri]

I’m researching some literature that isn’t really concrete, like each book defined a very different explanation of the “topic” and each of those topics in different books has different “characteristics”.

While going through different books, it’s becoming hard to see a good relation in the graph chart about how different books defined different characteristics for a topic.

For example, imagine that in one book, there is a topic about apples, and for characteristics of apples, you can see that it’s good for the brain and stomach and bad for fingernails.

In contrast, another book talks about the apple and says that it’s good for hair, bad for the stomach, and does not cover fingernails, and says instead of apples, bananas are good for the stomach.

Now if I want to know what is good for the stomach or what banana is good for and bad, I don’t have any way to find them other than going through all the notes again. Another issue is that I can’t see what the majority of books are suggesting about apple.

Or when I look at apple, I can’t see what characteristic is related to it.

What I’m trying to do

I’m trying to establish a mental model between the source (book), “explanation” and “characteristic”.

What I did

• created one folder per book,
• for each book, I’ve re-created the topics and characteristics again (48 topics and around 50 characteristics), so for each book, I’m recreating like 100 notes, let’s call them book’s note
• I also created about 100 notes as the root’s note
• then in each book’s note, I create a link to root’s notes. in this way if I go to root note, I can see under which book this topic is repeated.

Though this is working, but I have an issue; I’m creating about 100 notes per book, then I’m linking each book’s note to the root’s note. This cluttered the graph view tremendously, and I wondered if there was a better way of doing that.

I considered using tags, but tags won’t let you connect things. For example, I can’t connect #apple to #stomach and it won’t show as a connected node in the graph view.

I appreciate it if you share how you manage these not-concrete topics while one topic could have many meanings and contradictory relations and definitions.

[I-d-as]

Although I understand the sentiment of your use case, I am not sure I completely understand exactly what roots are. Regardless, if tags aren’t accomplishing your needs, it might be useful to create an Apple note instead of a tag. You can link to or from the Apple note in order to keep track of the relevant information.

When using a Local Graph, it can sometimes be advantageous to toggle incoming and outgoing links to achieve the desired view. Also, adjusting the depth value can be useful.

Finally, I have to suggest Obsidian search queries as a useful tool. It is worth scouring the documentation. These can be embedded in notes but will not show on graph. The same is true for the Dataview and eventually the Datacore plugins.

Good luck!

[mh_amri]

Thank you for your reply,

root notes work as an anchor that links all the books to the topics and characteristics
instead of connecting book A to book B

I created root notes that contain all the topics and characteristics.
Then inside each book, I link to those root notes.

a folder structure look like this

root (all topics)/
├─ apple
├─ banana
book A/
├─ apple 
book B/
├─ apple
book C/
├─ banana

When I want to find a topic, I open that root note, let’s say apple, I look at the graph, and I see that the root-apple-note is connected to other apple notes; one of those is connected Book A, and another is connected to book B, so I understand that Book A and Book B both talked about apples.

btw, I tried to learn more about your suggestions, but unfortunately, I couldn’t find any link to any of those suggestions you mentioned. Would you guide me with some links for each of them, maybe I can learn a little more.

Thank you

[holroy]

I’ve not used the graph a lot, as I find it mostly convoluted and messy (even for my very small vaults), but I think you need to somehow make the links, fields or tags contain more information differentiating the various variants of your topics.

One way to do this, which I think should reflect in the graph view, is to use tags. Like if doing #apple/stomach if it’s about the nutritional aspect of it, or #apple/fruit for the “fruity” variant. And possibly using multiple tags related to various contexts, like #apple/good but #orange/bad (not sure why though… ).

Doing this would both reduce the notes as you wouldn’t duplicate all the “apple” notes, allow for searching for similar tagged entries, and connect entries within the graph view.

It could possibly be useful to sub-class each tag tree, like if you define your main categories that you want divide stuff into. Like for an apple a bad category structure could be: #fruit/apple, #nutrition/apple #company/apple, #shape/apple… #movie/apple ? The idea is for you to imaging what kind of different categories makes sense for you to place the topic within, and then use that as base, and specify down the tree for various items.

Which categories makes sense to you, is up to you decide though, and I’m sorry I can’t think really good examples. Still I think the concept of using sub-topics/categories into tags, is the way forward, as you then can by a quick look determine which variant of the apple you’re looking at.

I’ve talked about tags, and I think that’s the best variant, but using inline fields for categorisation is also an option, although that doesn’t reflect into the graph view. Using more explicit note names could use the same thinking, like doing “fruit-apple” or “company-apple” or other schemes to include more information to differentiate the various alternatives.

[pwflint]

So I make one note per book, which usually runs several hundred to several thousand words long, depending on the book. Then I run back through the book note and summarize key ideas in their own atomic note. These get moved to a relevant folder while all the book notes get archived in a “book notes” folder.

I’m not sure if organizing notes to make the graph pretty is how I would approach a new or different system, but I agree that 100 notes per source feels excessive. I try to concentrate the bulk of my notes on original ideas or interpretations, and keep one note for each piece of source material.

PWF

[I-d-as]

@mh_amri I am just seeing this now. Local Graph (a core plugin) is described in the Obsidian documentation at this web address https://help.obsidian.md/Plugins/Graph+view#Local+Graph. In terms of Search, you can visit https://help.obsidian.md/Plugins/Search which also describes embedding queries at the bottom.

Dataview community plugin repository is here: https://github.com/blacksmithgu/obsidian-dataview. The documentation is here: https://blacksmithgu.github.io/obsidian-dataview/. You will notice the option of using fields to store metadata key value pairs. There is more than one way to do it. But, for example, you might have a line in a note like the following:

root:: [[Apple]]

By doing it this way, as opposed to simply linking to the root note, you could generate a table/list of links isolating all the notes that have a link to the Apple root note as the value for the root field. You would just add the following query in a note:

```dataview
TABLE WHERE contains(root, [[Apple]])
```

And finally, the yet to be released Datacore community plugin repository is here: https://github.com/blacksmithgu/datacore. Hope this all helps some.

Good luck!