240 - My Information Diet & Projects - I expose myself to a wide variety of information because I find it to be very enjoyable, sometimes to my detriment (scatterbrain at times). Thought I’d include this note here of what I’m currently reading/listening to in bits and pieces.
Updated Last: 9/10/2020 - I haven’t been contributing much to this zettelkasten lately because I’ve been reading books that don’t have to do with knowledge work directly.
I’m really bad with books, as I like to start them and am slow to finish them, the big pile of books phenomenon.
- Positive Psychology 101 by Philip Watkins (textbook)
- Social Science: An Introduction to the Study of Society by Elgin F. Hunt (16th Edition, textbook)
- Small Places, Large Issues: An Introduction to Social and Cultural Anthropology by Thomas Hylland Eriksen (4th Edition, textbook)
- Why Buddhism is True: The Science and Philosophy of Meditation and Enlightenment by Robert Wright
- Until the End of Time: Mind, Matter, and Our Search for Meaning in an Evolving Universe by Brian Greene
- Too Much to Know: Managing Scholarly Information before the Modern Age by Ann M. Blair
- Grad School Essentials: A Crash Course in Scholarly Skills by Zachary Shore
- How to Write a Thesis by Umberto Eco
- Make it Stick: The Science of Successful Learning by Peter C. Brown
I have a love/hate relationship with podcasts. They make for great background listening when you are doing busy work or manual labor type activities. On the other hand I get a ton of book recommendations out of them, which is distracting. I don’t listen to all these podcasts all the time, but instead listen to an episode if it sounds interesting.
- Sean Carroll’s Mindscape: Science, Society, Philosophy, Culture
- The Ezra Klein Show
- In Our Time BBC
- The Knowledge Project with Shane Parish
- The Psychology Podcast with Scott Barry Kaufman
- Tides of History
- All Songs Considered (NPR)
- Hidden Forces Podcast
- Lex Fridman Podcast
- Various Topical News and Politics Podcasts
These are a lot of projects going on at once, so many of them do not get love for long stretches of time. But I have a special bookmark on my browser for them, so they are not forgotten. They are just really long term projects.
- Private Zettelkasten - where most of my notes and readings go
- Obsidian Wiki - wiki I’m developing within my private notes collection, that I will eventually combine this with and make public using Obsidian Publish
- Obsidian Zettelkasten - this public zettelkasten
- Tools for Thoughts Zettelkasten - also created an experimental zettelkasten on the tools for thoughts forum because being able to make “note chains” using entire discords posts is a good experiment.
- Simple Zettelkasten Guide - an attempt make a beginners guide to the zettelkasten
- Pizza, Pizza, Obsidian Zettelkasten Example - showing how you can develop a concept within a zettelkasten
241 - Zettelkasten Forum Posts I Like
242 - Unique Identifiers
UID Primary Purpose
The primary purpose of each unique identifier is the quite obvious. It serves as a way to point towards another piece of information. Whether it be point to a whole another note (e.g. see note Metacognition) or section of a note (e.g. see Metamemory section in Metacognition note). With a digital system you can create a clickable link to it (e.g. see 202006011558 Metacognition). The whole point of it is it allows you to pull up more information when needed.
The key here is that the UIDs is unique and you have some method of tracking titles that keep them unique. In Obsidian, it automatically scans your notes so you can’t create duplicate titles. Because time flows forward, generating a timestamp of that instant will also always create a unique identifier that won’t ever be duplicated.
A contrasting example of a identifier would be the #tags system within obsidian. A single tag can represent many different notes, so it is no long unique to the note. Unless you track your tags and make sure they are descriptive enough to be unique.
UID Secondary Purpose
All the different options also have a secondary purpose besides serving as a unique pointer to that note or section of information. Titles for example give you a quick idea of what the contents of the note will be. Time IDs will give you an idea when the note was created. Luhmann IDs allow you to pretty much infinitely expand on any section of information in a physical system. People also misunderstand them and use them to show hierarchical relations between notes (not judging it).
UID Option 1: Titles
Titles are the go to standard for note UIDs. They have the capability to be unique, give you a brief idea of the notes content, and make note retrieval easier. Typically we search for information through retrieval cues that take the form of words. So it makes sense to search for note titles in the form of words.
The problem with note titles is there is no easy way to organize them so they stay unique. You could have them all sorted by title alphabetically but then that will become a giant headache when your notes are heavily interlinked. You’ll be jumping all over the place to pull up a set of notes. That is why Luhmann IDs work so well. They can be forever expanded upon while also keeping notes relating to one topic near each other. All these problems go away with digital software , as the computer can easily search all your notes for a matching title. Because of clickable links and automatic searching, collecting a the relevant notes become easier.
UID Option 2: Time IDs
Time IDs are another way to create UIDs for your files. They are so popular because they can be easily generated (just look at a clock) and will always be unique. In addition people find the Time/Date aspect of it very helpful because they see their notes and topics in a temporal aspect. For example, you remember writing a bunch of notes on a topic in January of 2015. It becomes very easy to pull up all your notes from that time period. This strategy wouldn’t work well in a physical collection because notes that are thematically related are not together. You run into the same problem as note titles above.
Another way to approach time IDs to is the method employed by Zettlr and The Archive (I think, don’t have an apple computer to test it out). Zettlr allows you to either generate a Time ID in the title itself or anywhere in a document. Then when you go to search it just does a universal text search for that time ID and uses a ranking system to display the findings. This means showing notes with the time ID in the title first, then notes with it in the metadata, then notes that have it show up early in the note, followed by every other note that mentions it.
Left off: Other downsides to Time IDs?
UID Option 3: Sequential IDs
Sequential IDs are where you give each note a number and just a next highest number when adding a new note. This is the strategy that Zettelkasten^3 uses. Because it is all a digital system, using Luhmann IDs become unnecessary as you can easily pull up related notes through linking and the creation of custom note sequences. The main downside to sequential IDs is that they serve no secondary function. Zettelkasten solves the search problem (pulling up the note you want) through the use of tags (keywords) and related notes being connected. So if you want to find a note on “meta-memory” you’d either click on the meta-memory tag or find the most closely related note (psychology or metacognition) and click through the note links until you stumble upon it. The case for doing this instead of directly going to the note is it forces you to expose yourself to other notes which might induce an unexpected but useful insight.
UID Option 4: Luhmann IDs
Combining UID Types
Because UIDs are just characters/filenames, you can use multiple types of UIDs in a single file. I’ve seen people over on Zettelkasten De Forums who are championing the use of Time ID + Filename + Luhmann ID, so a note would be titled something like “202006011558 Metacognition 4d5a”. The purpose of doing this is it allows for one to use all the secondary functions/purposes. The alternative to this would be sticking to one title type, while embedding the other information in metadata at the top. An example of this is TiddlyWiki, where it allows you to see when a note was created and last modified.
Where I’m at with UIDs
I’m still trying to figure out what the best option is, as I only occasionally come back to this topic and think about it more. I only have so much time, so I’m trying not to spend too much on the question. So far I’ve settled on combining Time IDs with Titles, so in Obsidian this note is titled “202009141037 Unique Identifiers”. I’m still not entirely sure this is the best option as I lose the ability to have unmentioned backlinks. I’m hoping someday Obsidian will implement partial note title matches for unmentioned backlinks. The primary reason I am using both is it future proofs me if Obsidian doesn’t work out. This is the strategy that the folks over at The Archive use, and I tend to trust them because they’ve put a lot of time into creating a zettelkasten program. Take this with a grain of salt though because I haven’t entirely worked out why this is the preferred method. My current understanding is they do this because it serves as a redundancy in case something gets messed up with the note title. I will add a section after this if I gain further understanding.
@lizardmenfromspace, Thank you. I’ve been looking for just such a explanation of a UID/Title/Zettelkasten ‘connection’ as you have provided here. Well done.
244 - Useful Features of a PKM Programs
Reflecting on and brainstorming useful features of a PKM program that I don’t currently see out there. There are a lot of programs with features I don’t use. What would be something I’d find myself using a lot?
- Existing PKM Features I use a lot?
- Breakdown the PKM process than think about what tools could help facilitate along each step in my personal PKM process
- What PKM problems more generally speaking?
- File Explorer - I use heavily, one of the main ways I find notes.
- Search Tab - primarily use for finding notes I haven’t worked on in a while and don’t immediately find using the hashtag index
- Hashtag Index - I add hashtags to the end of notes in such a way that it creates a basic index, whereby I’m easily able to search if I have a web/section of notes
- Linked mentions - I haven’t used successfully, it is nice when going through the creative phase to see what concepts connect to the current one. But I only have used it so far as a sort of quick back button.
- Star Tab - I use the star tab to track notes that I want to come back to because they are unfinished when I’m burnout or too distracted.
- Edit Button - Placeholder
- Pin Button - Placeholder
- Foot Notes - Placeholder
- Wiki links - not only do I use the wiki links feature (when you type [[ ) to interlink my documents, but I also use it to quickly find and navigate to other notes. Instead of going to the search tool, I just try to create a link to the note I’m looking for within the current file.
- Unlinked Mentions - I like the idea of unlinked mentions a lot but it is useless in its current form because my notes are in the format of “zettelkasten ID + title” (e.g. 202009161510 Aphantasia), and when writing I don’t use “zettelkasten ID + title”, I’d just use the title “aphantasia” in a sentence.
- Quick Switcher - I always mean to but just haven’t gotten in the habit of doing so, maybe someday I will. I also just flip over to the search tab instead.
- Outline - I feel like one day I’m going to use this but haven’t really found the need to yet because I have generally avoided creating really long notes that would make it useful.
- Open Random Note - I have this one saved, as I plan to use it more in the future. It is nice for selecting random note for the combinng process in remixing. See also localized random notes.
- Open Today’s Note - Placeholder
- Link with Pane - Placeholder
- Split Vertically/Horizontally - Placeholder
- Reveal in File Explorer - Placeholder
- Open Local Graph View - Placeholder
- Open Backlinks - Placeholder
- Star - Placeholder
- Open Outline - Placeholder
- Create a New Zettelkasten Note - I do not use this button because I have it mapped to a hotkey which I use instead
- Graph View - Placeholder
- Purpose of PKM
- Building a Knowledge Network for School
- Building a Knowledge Network for Work/Writing
- Building a Knowledge Network for Personal Use
- Information Digestion
- Choose Material to Process
- Process Material
- Index Material
- Interconnect Material
- Collection Maintenance
- Connecting Loose Notes
- Looking for New Connections
- Creative Writing
- Looking for Critical Mass of Notes
- Following existing arguments
I haven’t actually done much creative writing besides trying to synthesize two notes and making connections. All my writing has been done towards creating notes, not outside pieces.
- Lack of Precision
- Lack of Connection
- Shallow Notes
245 - Zettelkasten as an Abstraction Machine
Another way to conceptualize a zettelkasten is as an abstraction machine, whereby you abstract ideas out of a text then use them to create new ones using the copy, transform, combine creative remix. Below, looking at it through the REMIX FRAMEWORK
Read recommended books to understand their ideas and copy the most important Ideas into your notes collection. While books are the primary form of information gathering because they can be the easiest to process (fluid pause/rewind ability), you can copy information from anywhere. I use podcasts as a common source of information for my notes collection.
Generalization : a process in which detail is ignored to reveal a deeper structure. The term overlaps with abstraction, conceptualization, inductive reasoning, modeling, theorization, categorization, conclusion, unification, colligation, de-concretization, pattern extraction, pattern separation, and more.
Abstraction in its main sense is a conceptual process where general rules and concepts are derived from the usage and classification of specific examples, literal (“real” or “concrete”) signifiers, first principles, or other methods.
You can either combine a set of ideas to reach a higher abstraction, which would be like following themes across many different fields. Or you can combine concepts and ideas to come up with new ones.
A basic example of this would be to write out the sales pitch for a television show you love in one sentence. Do this twice, than write a third sentence that is a combination of the previous two sentences.
An ingenious real life example of this is how Gunpei Yokoi combined “withered technology” to make Nintendo what it is today.
The zettelkasten is good for creating conceptual connections through copying, transforming, and combining. This is typically done by combining information you already know very well (your primary field of study) with novel ideas you get from reading broadly.
This presumes you have a primary field of study, which not everyone has. Especially someone who is first starting out (e.g. first year college student). When this happens to be the case, then a zettelkasten can be used for understanding through creating a private model of the field. Understanding in this case, is literally uncovering all the concepts of the field and how they are connected to each other.
This is good and bad. The model you end up building in your zettelkasten can be useful because the information serves as anchor points for new information that is related. But it is bad because the information loses its utility over time. For example, when first starting a job, having pictures of all your coworkers with their names on it would be useful. But overtime as you internalize everyone’s name, then the pictures/names become useless. The information that was once useful, now just takes up space.
246 - Localized Random Notes - I like the idea of using the random note feature to select two different notes than ask if there is a connection between them. If there is a connection, than you create a middle note, otherwise repeat the process.
This feels like it may be unproductive because you’d have to go through so many bad note pairs before getting to a good one. I wonder if instead it would be a better idea to probe for local connections.
So a localized random note button would allow you to select two random notes that are 3-4 steps away from each other. This means that there is more likely to be a connection between them (I think) that you haven’t discovered yet. Besides the connection used to select them in the first place. A more direct connection (e.g. 1 to 1 connection) instead of the existing indirect connection (3-4 middle notes between them).
Actually I do follow this approach quite regulary - and in less than 1 of 10 cases this is really unproductive for me… If you’d stop thinking in categories for a moment, this almost always triggers a thought in your mind that had not existed there before…
248 - Zettelkasten Frequently Asked Questions
- Autonomous Notes: why do you want to keep notes autonomous? Why is it important?
- Atomic Notes: why do you want to keep atomic? What purpose does this serve?
Linked Notes: why do you want to link notes to each other? What purpose does this serve?
- Allows you to create connections between many notes (Themes)
- Allows you to create connections between two notes (Unique Connections)
- Understanding Luhmann: what made Luhmann’s Zettelkasten different from other note systems?
- Reformulate Ideas - Why do you want to reformulate ideas.
- Notes - Why do you want to take notes in the first place? What purpose do they serve?
- What makes a Zettelkasten?
What are alternatives to Zettelkasten? - Wiki, Note Card System
- What are the limitations of a Wiki that a Zettelkasten does not face?
- What are the limitations of a General Card System (Ryan Holiday) that a Zettelkasten does not face?
- What re the limitations of a common place notebook
- What is the key principle? - Principle of development?
- Zettelkasten and the Four Note Functions
My Zettelkasten Approach
We use digital systems instead of paper based ones, so a zettelkasten isn’t going to be a recreation of Luhmann’s system or similar ones. Instead we want to understand the principles behind what Luhmann was doing and understand how his system differentiated from other common note taking methods at the time. Using that knowledge, we can then apply it to digital tools and think up future tools that haven’t been created.
Why take notes?
- Long Term Memory - Traditional Notes
- Short Term Memory - Literature Notes
- Working Memory - Notes as Sketchpad
Question of Note Organization
One of the more important questions, which the zettelkasten system is answering, is how should you organize your notes? The primary purpose of organization is to make note retrieval easier in the future, so that you don’t have to go through every single note to find the one you are looking for. Organization can serve the secondary purpose of connecting information. It can help show you thematic connections between notes you’ve forgotten about and help you build a mental map/structure of topics represented by your notes, in the same way that you have a mental map of your family structure (parents, siblings, grandparents, etc).
Organizing Notes to serve the Core Note Functions
Note Card Boxes
Common Place Notebook
Folder Based Note Taking
Principle of Atomicity
Principle of Autonomy
Principle of Reformulation
Principle of Connectivity
The Key Principle of Development
How should I think about Note Titles?
What to take Notes on?
What information is worth taking notes on?
What is the purpose of Linking?
And how should it inform our note taking process?
How to Process Information
- Look for existing notes
- Create a single note on the information
- Create a small network of notes on the information
- Reformulate the Information
- Link the Note to other Related Notes
249 - Network Analysis is a set of integrated techniques to depict relations among actors and to analyze the social structures that emerge from the recurrence of these relations (Science Direct).
Alt: Breaking down a topic into a network of smaller parts in order to gain a better understanding of it.
How can Graph View help one better understand their notes collection?
My current issues with graph view is that I don’t feel it really tells me anything new about my note collection or helps me understand it better. This could be user error, as I just don’t know how to do such analysis. It could also be the tools aren’t well developed well enough yet to gain meaningful (new & surprising) insights about my notes.
Network Analysis Examples
- Netbase Quid specializes in text-based data analysis. Quid software can read millions of documents (e.g. news articles, blog posts, company profiles, and patents) and offers insight by organizing that content visually.
- Kumu is a powerful visualization platform for mapping systems and better understanding relationships.
- Vary Link Thickness
- Vary Node Size
- Vary Node Colors
- Vary Node Proximity (Forces)
- Vary Link Shapes
- Vary Node Shapes
Relationships to Map
Types of Tags
- Visualize Thematic Relationships
- Visualize Relationships between Keywords
- Connections between Notes
- Number of connections a note has
- Direct Links (Wiki Links)
- Indirect Links (Unmentioned Links, Backlinks, Tags)
- Metadata I’m unaware of
Future Visualizations for Obsidian
- Heatmap for Tags
- Timeline for Zettelkasten Notes
- Arc Diagram - nodes on timeline, arc links are hyperlinks to other notes
My comment that you quoted was in relationship to other zettelkasten besides Luhmanns. I was well aware of Luhmanns online, but don’t know of any other online ones to compare it against. If you come across any, let me know.
251 - Link as Search vs. Link as Direct
Came across this Zettelkasten De forum post that put something to words that has been on my mind lately. I’m still not sure what I want to do for file naming.
[] Name of file- is considered link-as-search, as it searches for just the Time ID. It is what is used by The Archive.
[[202005091413 Name of file]]- is considered link-as-direct. Where it searches for both the Time ID and filename.
[[ Name of file ]]- is the removal of the time ID, preferring just the title. This has the advantage of being able to support free-linking and unmentioned links. The disadvantage of this is if name of file somehow gets changed and doesn’t properly update in the other files. Or if you program of choice doesn’t update linked files with new name.
Related Zettelkasten De discussion “Future proofing my Zettelkasten – file names and linking”.
(Sarcasm Alert) Well, that doesn’t leave much hope for me. I’m looking to YOU for the answer. No pressure though. For me, The Archive (forum.zettelkasten.de) and Obsidian are two planets in the same solar system. (Roam is Roam, a Galaxy far away). Each are in orbit around the same sun, Zettelkasten. And there seem to be as many understandings, interpretations and definitions of things (Titles, Links) as there are posters to threads. So if you’re not sure, I’m really…F’ed! So, do me a favor. Pick something! So I can move on. Until you change your mind…
253 - Direct vs. Multistep Linking
This is fundamentally a question of structure and asking how does one build a model. When linking information (note to note) it is good to ask whether information should be a single step away (direct link) or multiple steps away (multiple notes between two notes).
The way I think about it is to ask “in what context would I be using this current note and what other information could further enrich it”. For example, when looking at my psychology note I don’t really need to have a link to a study on memory I found interesting. It is beyond the immediate context even though it fall in the realm of psychology. Instead I’d want information that directly relates to the concept of psychology itself, such as:
- the structure of the field
- major themes of the field
- most influential thinkers
- how the field emerged
- how the field relates to other fields
Instead the note I have for a study on memory would be linked to the note I have on memory, which would be linked to a note on major themes in psychology, which in turn would be linked to my note on psychology.
The reason why I’ve never seen the point of a UID is that I want the name of file, the unique ID, and the wikilink all to be human-readable and intelligible. In addition, the more I use Obsidian, the more I realize how much power there is in its [[autocompletion]] of bidirectional links, because it introduces linking at the point of composition – not just when searching afterwards. It’s changing the way I think while writing, as I reach for distinctive, complex concepts that I’ve already been developing. This might seem as though it stifles creativity, but I find that it actually helps me to probe deeper, building on thoughts I’ve already had, if only by using the same phrases. If I’m stuck while writing, I just hit
[[ and type a few key words, and I’ll often be reminded of a related idea that I can integrate into what I’m writing.
So I like the thought of UID other than standard titles is because they provide secondary features that can be useful (see below). But how it is structured in Obsidian, means I lose out on the utility of autocompletion and backlinks. I’m working on an Obsidian Publish site in a separate vault, implementing just straight titles, and I’m starting to like it more.
The main issue with titles is you lose that redundancy and sometimes you don’t always know what to title a note. I am still processing and trying to decide on what I’d like to do long term.
Here is my writing on UIDs so far
May I ask what happens to this project ? Did you find the ROI to be not worth it ? Just wondering about the reasons.
@minhthanh3145 it is still happening. I’m currently transitioning all the notes over to Obsidian Publish. It is just taking a very long time because I’m busy with life and reprocessing all the notes. Whereby I rewrite some of them, expand on the notes that are just bullet points, and better interconnect all of them. On the website index, if you look at department 8 - knowledge work, it has how many notes of the roughly 250 here that I’ve processed.
I do work on it every week though, so it is happening!
Thank-you for sharing not just your ideas, but the thought process that went into it.
I do have a Mac, so I am now toying with the idea of trying ZK3 and using Mac’s advanced multiple file renaming. But after checking my file count I am over 2 million created records on it, so it may not be practical.
I would like to hear mac users’ experience with practicalities and usefulness of file renaming.
Also, I wonder if macOS has
any file management features that assist in converting to a quasi ZK document management system.
You really got my brain working after reading this thread.