Zettelkasten Forum

Zettels sans title to mimic analog

Hi all,

New here! Been lurking for a bit, been making notes a lot longer. Super quick rundown, as it might be relevant to my question (feel free to skip and jump to the bold sections to tl;dr it):

  • I've carried a small pocket notebook on me for as long as I can remember, using a numbering convention of XXX.YY.ZZ where X is the "volume" number, Y is the page, and Z is the entry on the page, e.g., 092.04.01
  • I previously used Standard Notes, creating "discrete" markdown notes that basically just lived on their own (SN can't link); had probably 2 or 3k notes at this point just using that analog numbering system in digital form. My biggest issue, which is now obvious, was the "set it and forget" it nature of unlinked notes, making it a "stale" system.
  • Semi-recently, learned about Zettelkasten and the myriad connected realms, dove in, found linking to be extremely helpful.
  • Went overboard, used Obsidian, and went crazy with abusing automatic backlinks. And automatic backlinks with thousands of notes just get you "tag salad".
  • Then, finding options in The Archive and also iA Writer (testing in progress for both) after [very recently] switching to a mac has been great. Writing-focused. Links. Forced thought on backlinks. Incredibly helpful.

So, finally to the question I'm posing to you all, which follows from this tangential thread, and is something I've been noodling lately:

Do you think using a title-less filename would mirror some of the advantages of an analog slip box system? i.e., the note filename is just the UID (any convention, my particular one is not important).

I am a huge believer in analog writing for brainstorming, memory, etc, and I think this is undisputed over purely digital. I am, however, firmly in the "digital end point" camp, and will stay there with my text files (sorry! @scottscheper). However, in watching one of your video's Scott, specifically here @5:14, you mention the process of having to recall where to put something, thinking through related notes when going analog.

I think this digital, but sans-title might be a use case for someone wanting to use a digital system, but replicate some of the analog processing. My initial testing seems to confirm this idea, at the cost of a bit of friction.

The obvious rebuttal, which I see and have not made up my mind on yet, is the slower search. I will add though: you absolutely still can find a note that you have in mind quickly (for me, up to at least 5k notes, with a TON of notes still in sub-ZK-par shape mind you). It seems daunting, but even with a duplicate set of my notes with randomized numeric filenames, it took me just a few seconds to find a specific note I had in mind. The caveat I will grant is that I have a lot of notes, but about many different topics, so keyword searching in this way might be less effective for more narrowly focused zettelkasten (is that plural? haha).

However, there's enough friction that the first impulse is to use more easily-defined jumping off points in either structure notes, recent notes, or specific language/tags that force you to engage in this "analog-lite" way. Still more for me to test, before zetteling (not sorry!) on this reformat, but I'm curious what anyone else thinks!



  • @oddlynx Welcome to the forum!

    I'm not sure I fully understand or appreciate your question. It sounds like you are over-thinking the file name issue, though.

    I started with ZK about 2 years ago and have just been following the standard The Archive approach. But I rarely look at my list of files or file names when searching for materials. If I want to access a particular topic, I have 3 options - start with a structure note (if I happen to have one on that topic - not likely as I don't use them a lot), check to see if I have a tag that accesses that topic, or just use a simple search from within The Archive, and then once I've found a zettel of interest, follow whatever trail(s) of links that I have established between zettels. I use the last two methods about 40%/40% of the time and the first about 20% of the time.

    I am a believer in having a good title for each Zettel, a short version of which gets incorporated into the filename, (I use one of @Will 's macros for creating zettels). But that's about the only time I pay a lot of attention to the filename.

    By the way, I also use iA Writer on my Mac but on my iPad I find 1Writer works better for quick access to my ZK files.

  • Thanks @GeoEng51! And one question, and a clarification:

    How do you handle renaming within iA writer?

    And the clarification: I don't remember exactly what prompted the thought to just use the unique ID for the filename, but when I did test it out, I found that it reminded me a lot of the purely physical notecard/notebook style I used waaay back before ZK (just taking notes, no real system to it).

    I'd wager I'm not overthinking it though. I am quite familiar with the ZK system and have used both UID + title and just unique title for filenames to good effect for a while in various iterations. I'm pretty clear-eyed about the advantages of each.

    I am in the process of cleaning up my notes though, and in doing some more digging here on zk.de + my own brainstorming, I stumbled on the idea to make the system a bit more "opaque," which does seem to help me replicate the analog process to an extent.

    I find that as fast as the Archive and iA writer are, keyword searching for the title often makes me a bit too tunnel-vision-y, and reduces the "fabrication of serendipity" a bit for me. That was one of the things I was trying to fine-tune during this update.

    And! Maybe it's hogwash and it just works well with how I think, or it's just fresh and I haven't tested it sufficiently yet, but there were a few scant traces I uncovered during my delving in the forums here that I wanted to explore it further.

  • I could clarify: the "fabrication of serendipity" link probably isn't anything most aren't already familiar with. I just like the phrase. Kind of like @Taylor's "engineering serendipity" from this thread.

  • The desire to give a title to something is very human. The computer doesn't care, data is data, and strictly speaking, if your link and tag well, the title of a note does not matter. Yet, many of us feel different, see the discussions on appending the UID before or after the title.

    Personally I use titles and when I link to a note I many times link as [[UID Title]] instead of only [[UID]]. Apparently the title carries some meaning beyond the UID. But then again I find myself often linking and tagging to such an extent that the title becomes somewhat superfluous Example: a note with title "Metaphor of a Business Ecosystem" gets tagged as #metaphor and #business_ecosystem and linked to hub or structure notes on those topics.

    Sometimes a problem with titles is that I tend to write them before writing the content, thereby forcing the content in a certain direction. Example: a text in my writing ZK with title "UID romantic afternoon in the garden", might want to develop itself into something else, but I have limited my imagination by the title. I also find that when understanding on a topic increases, and my notes become more granular, the original titles do not always make sense anymore. Example: the "Metaphor of a Business Ecosystem" note might have been an atomistic thought a few years back, but it now appears that there are multiple sub-aspects, which make re-naming an re-arranging necessary.

    As a side-note, fabrication of serendipity is easy is you just create chaos, e.g, by linking everything to everything. Even with limited numbers of notes the permutations are endless. The price to pay is that many of those permutation are nonsensical, so there is a huge "search cost" involved in selecting the useful ones. There is a trade-off.

  • Thanks @erikh for the reply! And I've definitely had to rename and reorganize notes as they decay later on. I had't considered the title cementing the "destiny" of the note to some extent though by setting the title early on. That's an interesting angle worth keeping in mind. :smile:

    I guess I'm to blame here though. I should have kept my post more to the point. I'm not terribly interesting in debating the merits of a title-less filename. I think that's fairly well-defined and obvious. It's a design decision, with trade-offs either way. Likely more positives toward the standard convention, but a design choice nonetheless.

    Specifically, I'm just wondering if anyone has any thoughts on this aspect of my original post:

    Do you think using a title-less filename would mirror some of the advantages of an analog slip box system? i.e., the note filename is just the UID (any convention, my particular one is not important).

  • @oddlynx : sorry, as one more off-topic remark: The princes of Serendip are artful creatures. Reading your earlier post I suddenly realize that the fabrication of serendipity is directly related to one my major topics of study (exploration-exploitation dilemma) but I had never made a note on serendipity. I have now. Thank you for your serendipitous comment!

    On-topic: short answer: no :) But I think most analogue card or note systems have titles as well, right? Even the Dewey system has words next to the ID. About Luhmann I'm not sure. I think I have seen Zettel with and without title. Perhaps a Luhmann expert could chip in.

  • @erikh glad to help! And I just mean the filename having only UID. Obviously inside the note will have a title/explanation under the usual # UID Title or # Title convention.

    I'm finding in my initial testing that having to thumb through a few cards or traverse some sort of linkage feels a lot like having to actually look at a physical card.

    Of course you can thumb through cards and just look at the header, but that seems quite analogous to having to tab through a few files/links and peek at the header = title/description.

  • I can only guess, I have no experience with title-less approaches.

    I did try to hide the note list in The Archive, effectively hiding the search results and the file names. So all I had left was a search and the up/down arrow keys to "thumb" through results.

    That was weird.

    It got a bit better when I added a master super index note of sorts. A filename that would always be sorted on top thanks to starting with special characters like ___. I could access that via ESC + Down, i.e. clear search and select topmost note.

    That note was helpful because it offered a manual index of things I wanted to go to. With search alone, I could only narrow down results to some structure notes, but often times not to just 1 note, so I still had to thumb through results to get to where I actually wanted to go.

    So with that in mind, I'd guess that:

    • it is absolutely possible to ditch filenames
    • it requires either more thumbing, or more hand-picked navigational hubs

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

  • Thanks @ctietze! Yeah, it definitely takes a little getting used to. Though, I will say that it's easier if you can see how many notes you've filtered down to, even if the filename isn't contextually meaningful.

    I think the thumbing through the list reminds me most of analog systems. I seem to get new angles for an idea, or am reminded of tangential things to consider by the time I find the note I was looking for.

    I too went the master index route. I don't think this setup is doable without something more structured on the front end, e.g., similar @scottscheper setup. I also found myself adding a #track tag for the few notes that I have that are frequent use that are more pure utility, e.g., checklists for engineering revisions on drawings, purchase logs, etc.

    On finding a particular note, it seems to do just fine. What I'm not convinced about yet, is adding new notes before a structure is set up. It seems a little clunkier there, but it could also be me just not used to a new workflow.

    Anyway, thanks again for the input and testing it out @ctietze! :smiley:

  • @oddlynx said:
    How do you handle renaming within iA writer?

    I only rename files from within The Archive. That way, I keep the file name and the title in sync.

    More importantly, I often update the content of the zettels (they start off raw, as a seed, and then mature with time). When I do, I may update the title and the filename if it makes sense to do so. Hence, I don't feel too restricted by either the title or filename - they are just suggestions or reminders for content, in the beginning, inviting further thought and maturation.

    I have several tags that I use to indicate the state of the zettel (new, in process, getting there, mature). When I have time, I search on one of those tags and do some work to progress immature zettels. This can happen over many months. The process itself benefits from its cyclic nature.

    I find that as fast as the Archive and iA writer are, keyword searching for the title often makes me a bit too tunnel-vision-y, and reduces the "fabrication of serendipity" a bit for me. That was one of the things I was trying to fine-tune during this update.

    I look at structure notes, indexes, tags and even searches as simply methods of accessing a zettel that might be of interest to a particular line of thought I am pursuing. But the real magic is in following the bread crumbs we have laid down as we linked zettels in the past - reading all those linked zettels and opening our minds in a serendipitous manner to ideas that we might think aren't related to our current line of enquiry, but could be.

    Linking zettels thus becomes the most important aspect of creating and maintaining our ZK. We need to do it as we create zettels. We also need to do it as we progress or mature zettels, because our thinking about concepts changes over time and further links may become evident. But as I said, the magic or serendipity comes when we later follow links through our ZK, when we are pursuing a related or perhaps even (apparently) unrelated line of thought, and happen upon something that changes our perception or way of thinking.

    So I also have a tag that indicates the state of linking for each zettel, and part of my ZK maintenance activity is to regularly check poorly linked zettels to see if new links have become apparent (from more recent work on the ZK) or are possible. This is the work that enriches or even makes possible the future serendipity of our ZK.

Sign In or Register to comment.