# Some Beginner Questions

Hi All
I recently discovered ZK . This seems to overcome most of the stumbling blocks in effective note-taking.
I know how the method works but there are a few things that i want clarification/tips on.

The first question is about search and retrieval from ZK. I have read some people do not use tags for search and retrieval purpose. how do they search their ZK effectively? I guess only by note titles or search by a keyword they have in mind? But the problem i see with only having a title is that one may not remember them or the title may not accurately capture the though of the note. Tags seems to work ok, but they have a problem of their own: How many tags to use and which tags to use that makes retrieval efficient and effective. Tags may be too broad or narrow too.How do you go about retrieval problem and which approach has stood the test of time for you?

I also read people use structure notes. I dont know what they are and how they are used for retrieval purposes. Please shed some light on them as well.

Second question: When do u think branching is most useful from a note e.g. A1 A2 A3?

Third question I have seen people dumping tons of sentences on a single note which seems to, though not necessarily, contradict atomicity/preciseness of a note. How many sentence should a note have?

These question may have been addressed earlier in the forum but i thought people might have changed their perspectives from the last time they posted about these questions and other beginners may not have to dig deep to find answers to these questions.

Thanks
Aaron

• I use a combination of tags, meaningful titles and a search function in my zettelkasten (using vim with some plugins and scripts). When I know I can easily find something via tags, I look at the tagbar; When I still have in mind some title that points to the right direction I search for that in the zettel index, and if neither of those is the case I do a full search of my zettelkasten. This has always worked well for me and makes me independent from any single measure. It also means I don't have to obsess over tagging everything all the time. It's a form of A/B testing essentially.

Oh, and I also use the notorious structure notes you mentioned, mostly for very defined aspects though. My zettelkasten is still quite young, and the only structure notes I have are
1) a structure note (zettel) that is basically a list of all books I have worked on in my zettelkasten notes. Each of the zettels that refer to a book (often 10-20 zettel) are tagged with the family name of the book's author, so if I open the link to a book in the structure note I get referred to the index of all zettels tagged with that author's name (in this video you can see at 1:08 what it looks like when I open a tag and get all tagged zettel as an index in return, though i am using the tagbar as an entrypoint here, not a structure note:
https://www.bitchute.com/video/wlFZJ8ltQTgg/ ). And again, if I forgot to add the author's name as a tag to some zettel, there are always the links inside the tagged zettel which will probably lead me to them. Antifragility at its best
2) structure notes on historical topics, e.g. ancient history. In that strucutre note I will have a rough overview about the topic and links to the zettel that deal with specific topics in detail, e.g.

STRUCTURE NOTE ON ANCIENT HISTORY
tags: structure-note ancient-history
Paleolithic era -> 000183
Mesolithic era -> 000923
Neolithic era -> 000098

Opening e.g. link 000183 will then show:

000183: Paleolithic era:
tags: ancient-history
... text on paleolithic era ...

As for branching and zettel length: it depends on the content, I would say. For historical knowledge which I as a historian often deal with, I don't think you need to branch too much.
When it comes to intellectual discourse revolving around competing ideas, philosophy and other more "dense", less "factual" content, I would branch much more often to ensure atomicity.

Let me tell you one final thing: two months ago, I wondered about the same questions you did and found it hard to wrap my mind around the individual aspects of a zettelkasten, also because I read so many contradicting ideas. What I did was I tried various tools for weeks, realizing that none of them did what I really wanted to get out of them. If I were a mac or windows user, I would have probably sticked with ConnectedText (windows) or TheArchive (mac). But I use linux, so there's none of those available. I know some basic shell scripting and I know my way around the vim editor though, so I just configured that for my own personal needs, and am more than happy with that, the more so because I can keep adding new technical features to my zettelkasten someday if I want to (currently I am working on a script that automatically formats each zettel for printing so that I can also keep a physical zettelkasten as a hardcopied backup and for better visualization).

• @lunario Thank you so much for detailed response. It really clears most of the fog.

• @arron said:

The first question is about search and retrieval from ZK. I have read some people do not use tags for search and retrieval purpose. how do they search their ZK effectively? I guess only by note titles or search by a keyword they have in mind? But the problem i see with only having a title is that one may not remember them or the title may not accurately capture the though of the note. Tags seems to work ok, but they have a problem of their own: How many tags to use and which tags to use that makes retrieval efficient and effective. Tags may be too broad or narrow too.How do you go about retrieval problem and which approach has stood the test of time for you?

I also read people use structure notes. I dont know what they are and how they are used for retrieval purposes. Please shed some light on them as well.

@sfast Has said elsewhere that tags don't scale and that they are only a small piece of the puzzle. He also said he uses tags but doesn't rely on them but uses structured notes to provide a sharper focus.

My understanding of this is that both tags and structured notes are variations of the same thing, creating context and connections. Tags have a very small space for context, only a few words or phrases. You manually place them in other notes that you want to be included with the search results you get when you invoke the search by clicking the tag link. Because of the way we use tags, it isn't effortless to glean the context you are looking for from the search results the tag link returns. They work great for lower volume Zetttelkastens, but once you reach critical mass and tagged say 891 zettels with #BestTagEver other than creating a corral for the cows, what else do you have. A useful Tag example might be #Inbox or #wip for Zettels that need more processing or #Favoorite for the zettels you love the most. Tags are a piece of the puzzle and have their place depending on how you use them.

Basically structured notes act as the top level entry points into your Zettelkasten. With the Uber Top Level Structured Note(s) being the highest level entry points. Again structured notes are just another piece of the puzzle.

How effective and useful links, tags, structured notes, save searches, search links, etc. are all depends on the conventions you use and the way you think about them.

When all else fails, The Archive has your back by providing full-text search. How well you can build search clauses depends on how well they will work for you.

See the Create Welcome Note in the Help menu and check out the Search Expressions section for all the search operators you can use.

I hope this wasn't too much detail and I hope it helps. Feel free to ask for additional information and clarification.

Best of luck on your Zettelkasten journey.

• @arron said:
The first question is about search and retrieval from ZK. I have read some people do not use tags for search and retrieval purpose. how do they search their ZK effectively? I guess only by note titles or search by a keyword they have in mind? But the problem i see with only having a title is that one may not remember them or the title may not accurately capture the though of the note. Tags seems to work ok, but they have a problem of their own: How many tags to use and which tags to use that makes retrieval efficient and effective. Tags may be too broad or narrow too.How do you go about retrieval problem and which approach has stood the test of time for you?

The application I use searches the entire zettelkasten, not just the titles but the body of each note also. It would be nice if a saved search would just search the titles. This makes titles less important but yet another way to infuse the zettelkasten with context.

@arron said:
I also read people use structure notes. I dont know what they are and how they are used for retrieval purposes. Please shed some light on them as well.

Watch @ctietze's "Range" youtube series and it might help.
I use a structure note for about every book I read. I also use them for classes and writing projects. They become a major entry point into the zettelkasten though not exclusuvely.

@arron said:
Second question: When do u think branching is most useful from a note e.g. A1 A2 A3?

I 'branch' whenever I have a new idea.

@arron said:
Third question I have seen people dumping tons of sentences on a single note which seems to, though not necessarily, contradict atomicity/preciseness of a note. How many sentence should a note have?

I'm more interested in capturing and exploring ideas, notions, memes than I am in counting sentences. A note should have as many sentences as it needs and not one more. Sometime that might just be a sentence fragment and sometimes 1000 words. I am reminded of the writer who apologized for sending a long letter by stating "If I had more time I'd have sent you a shorter letter." This points to the fact that spending more time with a note could sharpen and shorten it. Condensing is hard work.

Will Simpson
I'm a futzing, second-guessing, backtracking, compulsive oversharing, ZK-maniac, in other words, your typical zettelnant.
Research areas: Attention Horizon, Productive Procrastination, Dzogchen, Non-fiction Creative Writing, Cognitive Workload, Python, Data Science
kestrelcreek.com

• I am also trying to figure out tags and I'm interested in what people think about the comment in How to Take Smart Notes that topics should not be tags. Ahrens says the problem with tags is that people use them as an answer to the question "What category should this be in?" which focuses on where to store information, rather than using them to answer the question "in what context will I want to stumble across this note again?" which (he says) more properly focuses on how to retrieve information.

This view seems to go against the most common ways I see people talking about tags, which is in terms of large topic categories and subcategories, but then, as @MikeBraddock says, what is the use of a tag that shows you 900 notes all related to "nutrition"?

The other problem I see with using lots of tags is that I've tagged many sources over the years but I haven't been tagging things accurately and reliably enough to trust my tag searches. So when I use the tag filter "paleolithic" I always wonder if there are some relevant sources or notes that I want to see but just forgot to tag.

The two uses for tags that seem most useful to me are 1) workflow status, such as #need_source, or 2) writing projects, such as #WWIIpaper.

Can others describe why they think topic-based tags are/are not useful, or what other purpose the use tags for?

• @cobblepot I have also been pondering tags recently. I reread the following discussion

In that discussion:

@Vinho said:

That article makes an in-depth analysis of tags and when and how to use them and a history of tags. It inspired more questions then answers for me but it change the way I think about tags and how maybe to use them. I am still pondering how to apply some of the ideas presented by Forte’s in his article. It is quite a read.

The enigma with tags is they are easy to use and powerful but how do you harness that power.

@cobblepot said:
I am also trying to figure out tags and I'm interested in what people think about the comment in How to Take Smart Notes that topics should not be tags.

The power of full text search can give you the same access to to notes that topic tags provide. But it can also be used to narrow the focus by being proficient in building search expressions using AND OR NOT conditions and grouping with parentheses. This seems like a viable solution for topic searches which you can then add additional conditions to your search expression to narrow the focus, for example:

• nutrition AND plants

Might whittle the 900 notes down to 25 in the search results.

I guess my point is using tags for other purposes than board topics retrieval gives us more pieces to the puzzle. But then again the search expression could have been:

• #nutrition AND #plants

Giving a search results of Zettels that only contain both tags instead of both words. I think The Archive would give the same results either way. So many ways and options to accomplish things, I find it overwhelming sometimes.

Full text search and being proficient with building search expressions I think would help with this question:

@cobblepot said
The other problem I see with using lots of tags is that I've tagged many sources over the years but I haven't been tagging things accurately and reliably enough to trust my tag searches.

Could Full text search and not tagging eliminate the concern of inaccurate tagging by depending of the text of your notes to reflect topics better, maybe.

I suspect all of our Zettelkastens have typos, miss assigned tags, other errors and omissions. You correct them as you come across them. Better to relax and go with the flow.

Structured notes and searches and saved searches are deliberate methods for finding and creating connections and leaving bread crumbs for your future self. Topic tags seem more incidental, less deliberate, less focused.

• I try to apply Ahrens' approach to tagging, that is when I create a new Zettel I think about any additional keywords I may search for in the future where I would like to see this Zettel come up.

Strategically, I see tags as a low effort way of grouping notes. Eventually tags group too many notes and you will likely want to organize them with structure notes instead. But that is an evolutionary step.

Finally, I let go of the thought of completeness with regards to tags. I believe I also read in Ahrens' book that Luhmann only made sure to connect new Zettels with some others but didn't bother to find all related ones. Also his index didn't point at all the notes related to some term but only a few.

• @nistude said:
I believe I also read in Ahrens' book that Luhmann only made sure to connect new Zettels with some others but didn't bother to find all related ones. Also his index didn't point at all the notes related to some term but only a few.

... namely the "hubs" that would forward you to other Zettel for more details, aka structure notes to create overviews in the terminology of our writing here.

Author at Zettelkasten.de • https://christiantietze.de/

• edited February 2020

@lunario said:
I use a combination of tags, meaningful titles and a search function in my zettelkasten (using vim with some plugins and scripts). When I know I can easily find something via tags, I look at the tagbar; When I still have in mind some title that points to the right direction I search for that in the zettel index, and if neither of those is the case I do a full search of my zettelkasten. This has always worked well for me and makes me independent from any single measure. It also means I don't have to obsess over tagging everything all the time. It's a form of A/B testing essentially.

Oh, and I also use the notorious structure notes you mentioned, mostly for very defined aspects though. My zettelkasten is still quite young, and the only structure notes I have are
1) a structure note (zettel) that is basically a list of all books I have worked on in my zettelkasten notes. Each of the zettels that refer to a book (often 10-20 zettel) are tagged with the family name of the book's author, so if I open the link to a book in the structure note I get referred to the index of all zettels tagged with that author's name (in this video you can see at 1:08 what it looks like when I open a tag and get all tagged zettel as an index in return, though i am using the tagbar as an entrypoint here, not a structure note:
https://www.bitchute.com/video/wlFZJ8ltQTgg/ ). And again, if I forgot to add the author's name as a tag to some zettel, there are always the links inside the tagged zettel which will probably lead me to them. Antifragility at its best
2) structure notes on historical topics, e.g. ancient history. In that strucutre note I will have a rough overview about the topic and links to the zettel that deal with specific topics in detail, e.g.

STRUCTURE NOTE ON ANCIENT HISTORY
tags: structure-note ancient-history
Paleolithic era -> 000183
Mesolithic era -> 000923
Neolithic era -> 000098

Opening e.g. link 000183 will then show:

000183: Paleolithic era:
tags: ancient-history
... text on paleolithic era ...

As for branching and zettel length: it depends on the content, I would say. For historical knowledge which I as a historian often deal with, I don't think you need to branch too much.
When it comes to intellectual discourse revolving around competing ideas, philosophy and other more "dense", less "factual" content, I would branch much more often to ensure atomicity.

Let me tell you one final thing: two months ago, I wondered about the same questions you did and found it hard to wrap my mind around the individual aspects of a zettelkasten, also because I read so many contradicting ideas. What I did was I tried various tools for weeks, realizing that none of them did what I really wanted to get out of them. If I were a mac or windows user, I would have probably sticked with ConnectedText (windows) or TheArchive (mac). But I use linux, so there's none of those available. I know some basic shell scripting and I know my way around the vim editor though, so I just configured that for my own personal needs, and am more than happy with that, the more so because I can keep adding new technical features to my zettelkasten someday if I want to (currently I am working on a script that automatically formats each zettel for printing so that I can also keep a physical zettelkasten as a hardcopied backup and for better visualization).

Looks very interesting! Do you have a git repo?
Reminds me of https://github.com/Hipomenes/vim-ide

• @Otternaut said:

@lunario said:

...

Looks very interesting! Do you have a git repo?
Reminds me of https://github.com/Hipomenes/vim-ide

No, my code is far too clumsy to be publishable on github. It's honestly just a bunch of grep/sed commands in for/while loops, extremely inefficient for most people's expectations but fair enough for me. I put it together in an afternoon basically and never cared much about making it universally usable (no relative paths set, formatting patterns based on my own idionsyncracies). But if you know just a tad bit about how to use the shell and vim you'll be able to put something similar up yourself.

• @ctietze said:

@nistude said:
I believe I also read in Ahrens' book that Luhmann only made sure to connect new Zettels with some others but didn't bother to find all related ones. Also his index didn't point at all the notes related to some term but only a few.

... namely the "hubs" that would forward you to other Zettel for more details, aka structure notes to create overviews in the terminology of our writing here.

I understand the ZK concepts as aiming to reduce hierarchy and encourage flat networks. At the same time, I can't imagine a ZK with no hub notes. However, isn't a reliance on structured hub notes going to eventually lead to the same problems that category tags would otherwise lead to? You wonder exactly where in the structure you should put any particular note, you try to make the hubs comprehensive so you don't miss any links, you worry about putting links to a note in all of your hubs so it's always included, etc. If this doesn't happen, can you help me understand why?

Order put the question another way, I thought that Luhmann's index idea was that an index word had to link only to a note that would get him into the network of linked notes about that topic, not the index word had a link to a note that would itself act as a structured index of all notes relevant to that topic. I'm worried that reliance on structured notes as central organizing principles of the ZK will undermine the goal of preventing organizational paralysis whenever classifying a new note.

• I think it helps to turn the context of the question a bit: it's not about where does this belong?, like, "officially", but rather how do I want to access this later? That's evident when you link to a note from 3 structure notes that represent 3 different writing projects. The note is used in these because it serves a purpose, not because it needs to be there for some objective reason.

Author at Zettelkasten.de • https://christiantietze.de/

• I agree with your idea about focusing on access, but I'm not getting how that relates to my concern. Maybe I'm taking the term "structure" too literally. I've seen reference to structure notes on these boards to mean 1) lists of notes related to a topic, which raises the potential concern about completeness I mentioned above, and 2) a list of links organized to serve as the beginning of a writing draft.

Maybe an example would help. I create a "structure note" on the general topic of "interactivity". That topic relates to a lot of things, so as the list of related notes grows, to navigate it I divide it up into sections, and then I start worrying about whether I've divided it into the right sections, whether to put a new note link in one section or another (not both, because linking a note in every section creates the same mess the division into sections was supposed to relieve), etc. How would you approach dealing with this problem? Thanks for the comments.

• edited February 2020

@cobblepot said:

Can others describe why they think topic-based tags are/are not useful, or what other purpose the use tags for?

I use tags to identify sources for information in a note (last name, dat of pub. for books, last name, title identifiers,:date of pub. for articles: #barthes1957, #brewerMHEL2010, etc).

Started ZK 4.2018. "The path is at your feet, see? Now carry on."

• @cobblepot said:
Maybe an example would help. I create a "structure note" on the general topic of "interactivity". That topic relates to a lot of things, so as the list of related notes grows, to navigate it I divide it up into sections, ...

You start off on the right track till you get to here...

Then you wonder off concerned about the "right" way to do things, worrying about completeness. Do what feels good with what knowledge you have now and later when you look at the note again your knowledge will be different and you'll naturally wonder "What was I thinking?" This is how we know progress and learning are happening.

and then I start worrying about whether I've divided it into the right sections, whether to put a new note link in one section or another (not both, because linking a note in every section creates the same mess the division into sections was supposed to relieve), etc.

I'd be more worried about being paralyzed into inactivity because I was holding myself to some unattainable standard. Been there, done that, found it futile. Follow Voltaire's advice, "Don't let the perfect be the enemy of the good." If you get your notes 80% complete, that is 79% more than you would have by not creating a note that was "only" 80% correct/perfect/complete. I only want to get my notes up to maybe 25% complete when I write them. I don't worry because I know the I'll edit, revise, reference, add, split the note as time moves forward if the notion expressed in the note continues to tickle me. Otherwise, it will languish but I'm ok with that. I can't predict what notions I'll be working with next week let alone in 10 years. My knowledge grows, morphs, and I hold it as a dynamic thing and resist any efforts to make it static.

In my case, I use my Zettelkasten for molding my entire intellectual life, which is in constant flux. I might feel different if my use case was strictly a dissertation and a 100% meant a prestigious career.

How would you approach dealing with this problem?

Relax. Iterate don't ruminate. My first 500 notes where crap but they got me to where I am now. For that I'm grateful. You may progress faster than me, I'm notoriously slow.

Apologies for the preaching.

Will Simpson
I'm a futzing, second-guessing, backtracking, compulsive oversharing, ZK-maniac, in other words, your typical zettelnant.
Research areas: Attention Horizon, Productive Procrastination, Dzogchen, Non-fiction Creative Writing, Cognitive Workload, Python, Data Science
kestrelcreek.com

• @cobblepot said:
Maybe an example would help. I create a "structure note" on the general topic of "interactivity". That topic relates to a lot of things, so as the list of related notes grows, to navigate it I divide it up into sections, and then I start worrying about whether I've divided it into the right sections, whether to put a new note link in one section or another (not both, because linking a note in every section creates the same mess the division into sections was supposed to relieve), etc. How would you approach dealing with this problem? Thanks for the comments.

While totally on the same page as @Will with the 'do not sweat it', I do feel you as I am in more or less the same position of finding my way in a new methodology.

I am in the process of both transforming my current notes into a ZK format (previously on nvAlt with Douglas Barone's FSIM background - more on that another time), and also getting a piece of my extant physical notes entered into the ZK. I cannot get to the point to just ditch all the work I put into these, often chaotic and random, scribbles I made during my master's studies of philosophy (2014-2017). There is so much good stuff in there, if only it would be visibly interconnected and related, there is material for at least 10+ papers and possibly a book or two.

Anyhow, now on to your question, and the way I look at it.
It is not really a 'problem' but more like an inherent quality or feature of a ZK (at least the one I am implementing in The Archive). I'll explain with an example here.

I have many notes on John Dewey and his philosophy. One of those (atomic) notes is on how his notion of habit relates to the philosophical concept of immediacy.

In one of my structure notes on the topic of moral imagination, I talk about which constraints there are on a Deweyan moral imagination. One of those is related to our habits and how they make issues 'immediately clear' for us, in that we do not consider alternative explanations. In my structure note, I have thus a link to this atomic note on 'habit and immediacy', providing some context in the structure note as to why this is relevant for moral imagination.

In another of my structure notes, this time on intuitive moral judgments (this grew out of a paper of mine called 'Habit Based Reconstruction of Intuitive Judgments in Social Psychology: Dewey and Haidt on the Role of Intuition in Morality'), I have the same atomic note on 'habit and immediacy', this time linking it up in a different context. Here it provides the function of showing there is another way of thinking about 'immediacy' than in the standard philosophical dichotomies of intuitionist / rationalist debates, of which the work of Jonathan Haidt is the latest popular offshoot.

Bringing this slightly expansive digression back to your question, I would advise you to just link this atomic note up twice, even if in the same structure note, when it provides two different contextual links or connections. As long as your structure note makes clear why it provides these two different functions, there is never a real problem of meaning.

Of course, in a future iteration of your structure note, which is what @Will hints at too, you will always reconstruct. This could lead to an intermediate note (my idea: interactivity in organism and environment), in between your structure note on 'Interactivity' and this atomic note (my idea: interactivity interpersonal in speech), that is on the multiple ways 'interpersonal interactivity in speech' comes to the fore when you think about interactivity inbetween organism and environment. For example, one interpretation or explanation may be about 'talking is nudging others into taking your point of view'; the other might be 'talking is co-evolving a dialogical niche'.

Hopefully, this helps. Sorry if the examples are way out there maybe, but these topics are currently occupying me a lot.

I am a Zettler, ie 'one who zettles'
research: pragmatism, 4e cognitive science, metaphor | you can't be neutral on a moving train

• @Will said:
In my case, I use my Zettelkasten for molding my entire intellectual life, which is in constant flux. I might feel different if my use case was strictly a dissertation and a 100% meant a prestigious career.

No worries about preaching - I am in total agreement, generally, not to sweat the small stuff. Unfortunately, since I'm an academic on the tenure track and am using ZK for publishable research, 100% does, indeed, mean a prestigious (steady?) career.

I think I may have misled you by talking about putting things in the "right place". What I really meant was that I start to think, "where can I put this so that I'm confident I'll stumble across it again when I'm looking for it?"

• @John, thanks, that example was actually quite helpful (perhaps in part because I am familiar with the topics you referenced and those you list in your signature). I think part of the challenge for me is that, like you, I already have hundreds of notes and a lot of domain-specific knowledge in my mind, so the ZK right now is as much of a reorganization process than a knowledge-building one. If only I had discovered it years ago!

• @Phil said:
I use tags to identify sources for information in a note (last name, dat of pub. for books, last name, title identifiers,:date of pub. for articles: #barthes1957, #brewerMHEL2010, etc).

Thanks, I would consider these workflow tags (identifying sources) rather than topic-based--so it seems that you agree with my skepticism about topic tags.

• @cobblepot said:
Thanks, I would consider these workflow tags (identifying sources) rather than topic-based--so it seems that you agree with my skepticism about topic tags.

Yes. For the reasons stated upthread (challenges in tracking which tags I'm using for a given topic, redundancy with note titles, etc.), I'm still experimenting with different approaches to tag use in the ZK. When I do use a topic-based tag, I'll try and capture a broad and relatively frequently tackled idea which captures lots of different iterations in different contexts--#ambivalence is a good example, which tries to throw a broad outline around many different notes which focus on nonduality in literature, critical theory, Buddhism etc.

But in general, I'm still looking for a useful way of deploying tags.

Started ZK 4.2018. "The path is at your feet, see? Now carry on."

• I think for me, tags were (and still are) always something of a problem.
I keep forgetting how I tagged other materials, was it singular/plural, was it in capital or not, et cetera.

However, I do think there is a time and place for tagging. The way @Phil uses it for bibliography / citekey is something I have recently adopted as one extra way to navigate across my Zettelkasten.

And I think here lies the 'takeway' if you will. Use whatever way of tagging that works in your Zettelkasten. Tagging, and tags, are tools used in working and thinking with your system.
Therefore, it can only be that what others prescribe is a general principle that you always have to see how it works in your case.

For example, right now I still have some leftover tags from the methodology I used to work with on my notes archive (Douglas Barone's FSIM). In this method one of the ways to slice and dice through your notes archive was the axis 'tags', which were a system of single capital letters and a number (say, R1 or T2, standing respectively for Record 1: Diary entries, and Thoughts 2: Observations, non-fiction). It's quite abstract and arcane at first, but after working with it and adapting it to my tastes this started to make sense. More importantly, in any OS or application, files sorted first on date (yyyymmdd) and then on tags, and then a short unique filename (example of an existing file in my not-yet-transformed-to-ZK archive: 201608171240-N2-STUDY-Dewey-Experience_and_Nature-Chapter_5_Nature_Communication_and_Meaning - here you see I already added hour and minute to the datetime, and after the axis tag a topic identifier 'STUDY').

Whilst working in my Zettelkasten last week, and reading up here in some discussions on the forums, I started thinking that I should not worry too much about all these leftover vestiges of an old system. I still 'think' a bit in the FSIM methodology (mostly the axis tags), and in a sense this gives me an extra filter or way of building searches for the omnibar to find certain specific types of notes (for example, in my old system any N6 was a verbatim quoted passage from a source without any thinking of my own added).

[TL;DR] I suggest thinking about the ways you intend to use your ZK. If you have certain types of notes that you want to be able to filter out based on criteria (for example what FSIM allowed is by stage in the writing process from Notes to Thoughts to Writings) then make up those tags, and it is powerful to then play around with in your Saved Searches to be able to slice and dice up your ZK in that way.

I am a Zettler, ie 'one who zettles'
research: pragmatism, 4e cognitive science, metaphor | you can't be neutral on a moving train

• And to add more to the tag talk.

I think the Omnibar is the answer to worrying about tags and there ability to scale, as well as case concerns, and maybe any other ill we think tags might have.

### Omnibar and Search

1. The Omnibar. The Omnibar commands the list of results. It filters all the files in your archive via full text search. -From the Welcome to The Archive help note

Keywords there are filter and full-text search. The Omnibar has search expressions you can use to do complex filtering.

#### Search Expressions

The Omnibar performs simple text search, but also offers boolean search expressions and other search niceties.

For example, when you surround a phrase in quotation marks, "like this", it will search for the phrase, and not the single words it's made of.

Then there are logical operators:

• Operators are all caps: AND, OR, and NOT.
• foo bar baz is equivalent to foo AND bar AND baz.
• Instead of writing NOT, you can prefix a word or phrase or parens with !: NOT b is !b and will show notes that have no "b".
• Phrases win over the exclamation mark: "!b" is a phrase search for "!b", matching the literal exclamation mark.
• Instead of surrounding something with quotation marks, escaping with a backslash works, too: \!b will treat the exclamation mark as part of the search string.
• Escaping quotation marks inside phrase searches will also work: hello "you \"lovely\" specimen"
• You can parenthesize expressions to group them. For example, !(foo OR (baz AND !bar) is equivalent to writing !foo OR !baz AND !foo OR !bar.

A real-world example: "the archive" OR # zettelkasten will not match the and archive alone, as you'd expect from phrase searches. You will get all notes that are tagged with # zettelkasten, contain the phrase "the archive", or both.

If you're into algorithms, please note that OR and AND do not have correct operator precedence; I found that doesn't matter for searching notes. So if in doubt, better use parens to group parts of the search! -From the Welcome to The Archive help note

By using AND OR and NOT and parentheses for grouping I suspect you can find anything. Tremendous power there and probably The Archive's most killer feature. This allows us to have tags that don't really scale but they still can be useful for narrowing our search results by using search expressions.

For example:

Do I have notes tagged about atomicity or structure not related to zettelkasten?

(#atomcity OR #structure) NOT #zettelkasten

Can also do this as a SEARCH link inside a note:

[[(#atomcity OR #structure) NOT #zettelkasten]]


### Case sensitivity

All of these yield the same results except for the typo Zkmethodx. The Archive ignores case while searching. No need to worry about string case when throwing tags around.

There are two KM macros for doing full text find and replace that can help with the typos. Thank goodness because I am the king of typos.

Here for title text find and replace: https://forum.zettelkasten.de/discussion/comment/3064/#Comment_3064

And here for full text find and replace helpful for tags: https://forum.zettelkasten.de/discussion/comment/3084/#Comment_3084

I would add the tag cloud KM macro for visibility into what tags you are actually using in your Zettelkasten.: https://forum.zettelkasten.de/discussion/comment/2505/#Comment_2505

### Final Thoughts

My original plan was to figure tags out before going and deep and gungho with The Archive.
Tags can be a good (we don't need perfect) and vital piece of the puzzle with the current state of our toolkit.

I think I am just going to let the loose end drag and be happy based on this discussion.

My approach is varied. I don't put much effort into tagging as for the most part I haven't seen nor have I been convinced of their usefulness. I usually don't apply them when I'm composing the note and add them later as I find connections and want to group or record connections with other notes. Sometimes weeks or months after the note origination.

One area where I have found some use is tagging projects. I have too many projects going on but for example, I'm writing an essay on wolf reintroduction to the Hells Canyon Wilderness. I have notes where I'm processing research papers, have first drafts of sections and scenes, and a structure note with links into my broader Zettelkasten, all tagged with #High_Trail. This will facilitate exporting the notes together for the second draft work which will start soon.

This is a different example of how I use a structure note. I haven't yet got sophisticated enough to have a structure note linked with another structure note, like to galaxies merging it would likely set off a tremendous spectacle. It will happen eventually, I hope we're prepared.

Will Simpson
I'm a futzing, second-guessing, backtracking, compulsive oversharing, ZK-maniac, in other words, your typical zettelnant.
Research areas: Attention Horizon, Productive Procrastination, Dzogchen, Non-fiction Creative Writing, Cognitive Workload, Python, Data Science
kestrelcreek.com

• This is a very enlightening discussion. Thank you everybody. I will ask more questions ,if needed, after reading all the comments.

• @Will said:

Hah indeed.
Today I stumbled upon a command-line tool that specifically works for the Mac:
Tag: A command line tool to manipulate tags on Mac OS X files, and to query for files with those tags.

I have just forked this code to see, as a hobby project, whether I could get it specifically working in a Zettelkasten setup. To direct it at my notes so to speak and manipulate them or query them.
I mainly think in a way alike the Keyboard Maestro scripts that generate tag cloud overviews, to get a feel of your most used or underused tags, but also to easily check and correct spelling mistakes or to find/replace tag x for tag y.

I am a Zettler, ie 'one who zettles'
research: pragmatism, 4e cognitive science, metaphor | you can't be neutral on a moving train

• @John said:
I mainly think in a way alike the Keyboard Maestro scripts that generate tag cloud overviews, to get a feel of your most used or underused tags, but also to easily check and correct spelling mistakes or to find/replace tag x for tag y.

Because we can store the "Tag Cloud" as a variable in Keyboard Maestro we then can call that variable to check our spelling. I use a macro that provides a list of already established tags from which to choose from. The macro then pastes it into the note. You can choose multiple tags or ESC out of the macro and add a new tag. This prevents the problem of #metaphor and #metaphors tags or in my case #high-trail and #High_Trail.

A rough gif showing the macro in action.

Keyboard Maestro - Insert Tag

Will Simpson
I'm a futzing, second-guessing, backtracking, compulsive oversharing, ZK-maniac, in other words, your typical zettelnant.
Research areas: Attention Horizon, Productive Procrastination, Dzogchen, Non-fiction Creative Writing, Cognitive Workload, Python, Data Science
kestrelcreek.com