Zettelkasten Forum

Projects Separated into Different Folders. Bad Idea?


I just joined. Currently reading How to Take Smart Notes, by Sonke Ahrens. I can definitely resonate with the concept, and I have been experimenting with both Sublime-zk and Sublimeless-zk. Have been also working to sync the corresponding User file between my Windows 10 box and my MacBook. So far, so good.

Where I run into trouble for my intended usage--cross-connecting ideas derived from different books--is where I put my Zettels (notes) from different book readings into separate Project folders. When I do that I cannot link reference notes from other such projects (archives).

It seems that the organizing idea must be to throw everything into one so-called archive. But then my notes to a specific book get lost in a sea of other notes.

It would be nice if I could be in one archive and link-relate to Zettels in another specifically organized archive.

Am I missing something?



  • Welcome, @GreenBeing! Your first hunch is quite normal, I found when I tried to get students at Uni to adopt the Zettelkasten Method: to keep things separate and organize with folders comes rather naturally. And you also see the way out: use one single archive for all your work.

    Mixing notes is not as bad as you make it sound. As long as you have a full-text search available (and that is always when you're in front of a computer), you can utilize the search to filter all notes down to all notes on book X. I put book citations into my notes, so I can search for the book's citation key. Rene's tools support BibTeX citations natively, too, so I suggest you use that!

    After a while, you'll find that you won't be looking for "all my notes from this book" anymore and think in subjects instead. That's where tagging kicks in. You get a cluster of notes related to a topic when you search for a tag like #zettelkasten (e.g. producing not only notes on Söhnke's book, but also those about this blog, forum discussions, Nicolas Cifuentes-Goodbody's YouTube videos, and so on).

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

  • Hi @GreenBeing,

    have a look into the concept of structure notes:


    I currently have 7400 notes based on many books and articles from various fields in one folder with not problems at all. A good system scales with your thinking organically or it isn't good.

    I am a Zettler

  • Hi @ctietze, @Sascha and thanks for the responses. Sorry for the late reply; I was traveling.

    Right after posting my original post to this thread, I did a search in the Forum concerning structure, projects, and folders. I almost immediately came across @ctietze's "emerging-overview-with-structure-notes" post suggested by @Sascha. Sascha's "A Tale of Complexity – Structural Layers in Note Taking," also suggested by @Sascha, makes a very good companion post.

    Moreover, after reading "How to Take Smart Notes," by Söhnke Ahrens, I am convinced of the need to abandon "folders" as an organizing schema and use a single, scalable archive that is supported with tags, citations, and a powerful search engine.

    However, this makes me wonder about the utility of Rene's Project function in his implementations of sublime-zk and sublimeless-zk. How is it intended that we use this function? To be sure, though, using Rene's BibTeX citations should indeed obviate the need for book-oriented folders ... and projects.

    Ahrens suggest the smart use of tags. I would be interested in hearing how others approach the system of tagging. My first hunch is that the field of tags used should be a semantic mapping into one's knowledge interests ... kind of like book titles found on one's bookshelf but condensed down to single-word tags, perhaps like #cosmology, #biological_evolution, #earth_science, etc.

    I am rather glad that I didn't get too far with my original schema before asking for some advice. :)

  • Tags don't scale. So, if you plan to build a life long system, don't rely on them. It is not a viable long-term solution to emphasise them.

    I am a Zettler

  • @Sascha said:
    Tags don't scale. So, if you plan to build a lifelong system, don't rely on them. It is not a viable long-term solution to emphasize them.

    @Sascha, thanks for weighing in! Because of your Avatar, I sometimes think that I am talking with Daniel Quinn's Ishmael, and that's a good thing. :p

    Still, your response about tags is somewhat unexpected because there is so much discussion on this forum about ways to use tags, say, to create structure, searching for content, creating tag hierarchies, etc. [and, I can't seem to resolve what the double hashtag tag does over a single hashtag. Can someone please enligthen me? Thanks!].

    But, I think that I understand where you may be coming from; the underlying semantics of a person's tagging system can change over time leading to inconsistencies, and there is no good way to go back and correct for that. Of course, in social networks, this can be a big problem: "A word is worth a thousand pictures" kind of thing. But when we are talking about a single user's tagging behavior, perhaps there is less danger in that. Not sure.

    And all of this is why I raised the question to the forum. Ahrens advises smart tagging (I confess that I cannot find where he said this. :( ). So what is that? I'm still reading and researching; so maybe I will find out. My sense of it, though, is that you have to be careful with the semantics and keep the tag general enought to provide a reliable way to find an entry point into a collection of linked Zettels that will serve your immediate purpose ... kind of like keywords once used for SEO [actually search engines no longer use these because they are gamed ... leading to poor user experience].

    Niklas Luhmann developed what Ahrens calls his index. He used this to find one or two Zettles as an entry point into a line of thought or topic. Zettels with a sorted collection of links would be best. I am assuming that this index is some kind of a tag architecture. Not sure ... yet. However, this method would suggest that tags only be used on Zettles that can serve as an entry point into a line of thought or topic. These Zettels would be characterized by sorted lists of links, perhaps in the form of a TOC.

    [BTW, Ahrens also talks about keeping your literature notes [[project notes?]] separate and with your reference system perhaps pointing to project folders. What goes into your Zettelkasten is only a distilation of these notes in terms of insights you are working on. So, it would seem that our referencing system would need to accomodate such a project-oriented archive. I talk about reference managment systems in another thread. Of course, all Zettles--in the Zettelkasten--based on literature notes would contain bibliographical citings.]

    In thinking about this more, I can see the use of tags as standing in place of categories. By that, I mean categories can be thought of as sorting mechanisms that would effectively clump Zettels into very general groups ... kind of like the ones I suggested in my last post in this thread: #cosmology, #biological_evolution, #earth_science, etc. But used in this way, each Zettel would be required to be categorized. Tags are more optional. So, I can see some danger in that.

    But, if I were to guess what your solution would be to implement a recall-grouping strategy that does scale, I think it would be links. Certainly, manually linking Zettels is much more holistic and deliberate in associating Zettels.

    I will try to say more as I get through How to Take Smart Notes. Doing multiple things at once.

  • I'll have to write a blogpost about this.

    In short: If you have a big archive you'll get a lot of search results from many tags. If you want to connect notes via tags the more notes you mark with one tag the blurrier the connection will become. I didn't thought about scaling until I was at a point of having a bigger archive (for me probably around 4000 notes).

    Techniques shouldn't give you different results no matter how big your archive is. If your archive behaves like a complex network that should be more complex than you to understand (why bother with a second brain if it is just a sophisticated filing system?) you should employ methods that are designed to deal with overwhelming complexity.

    1: Hartmut Freyhof, Hans Gruber, and Albert Ziegler (1992): Expertise and hierarchical knowledge representation in chess, Psychological Research 1, 1992, Vol. 54, S. 32--37.

    2: Long-Term Working memory: https://www.ncbi.nlm.nih.gov/pubmed/7740089

    I came up double-hash tags to mark core structure notes. Only the central structure not about xyzy would be tagged with a ##xyzy.

    The index had no architecture to it. It was alphabetical. Luhmann relied mostly on direct linking with various techniques (even Folgezettel can be mapped as direct links).

    This is derived from actually using my archive. It is more like practise explained then theory put into practice.

    I am a Zettler

  • Hmmm… I worry now that I have misunderstood something both important and basic: I did not expect @Sascha to critique tags in the way he does above. Until this moment, tags have seemed a vital element of creating notes.

    @Sascha, are you saying you no longer add tags to notes (in the hopes of them turning up in future searches?

    Using ## to denote header topics seemed brilliant to me to get to an index of that topic vs the topic itself. But I somehow feel that if that’s all there is, then there is (almost?) no such thing as creating a quick note to capture a passing thought. Once I capture the thought, am I then to spend hours laboriously searching and linking all notes to all other notes on the topics and then linking them also to the ##indices?

    I should extract this and make it a separate note, but I confess I haven’t yet really started a Zettelkasten, because I’m afraid that I’ll mess it up (and make it unusable) before I even start.

    Part of the reason is trying to figure out all the codes everyone seems to be using in titles (?) or near the top of the note. I didn’t expect to confess this now, so I don’t have any examples, but I think I remember a part of one @Sascha uses: Ü…. For outline?

    I also seem to recall that these codes have to do with some suggestions on codes suggested in the Scrivener forum based on a couple posts by a really smart user (can’t recall the name).

    Now in addition to not starting because I don’t know all the codes (is there a post somewhere?), I’m concerned about tagging or not tagging. What a mess I seem to have created! ????

    I clearly need to reread @Sascha’s article on the ##index!

    Apologies for angst-y post!


    fabricando fabri fimus

  • Re the hub cards, doesn’t a search for #motion also return ##motion? Perhaps there are so few ## hits, that it’s easy to ignore them while scanning the results?

    fabricando fabri fimus

  • No, I still add them and use them to specify some searches. But they are not good organisational tool for big archives.

    I do quick entries like that: I search a structure note that has something to do with my idea. I branch from there (with The Archive via the wiki link thing) and write as my thought develops. Then I do a couple of links that trigger more thoughts. But I don't go through all notes and try to squeeze every possible link in I can come up with.

    "Ü" is for "Überblick" (German for "Overview") But it is arbitrary.

    Perhaps there are so few ## hits, that it’s easy to ignore them while scanning the results?

    Yes. And you want them included because you are searching for all notes related to this keyword. The structure note on that topic is clearly related to this topic. :smile:

    Just get started and learn by doing. There is not much value in hesitating. :smiley:

    I am a Zettler

  • fabricando fabri fimus

    fabricando fabri fimus

Sign In or Register to comment.