Zettelkasten Forum


Link Discipline --- only link to older notes?

I'm curious, do those of you who are mindful with your notes and links make sure to only link to "previous" notes (those with a lower ID) when creating a link.

Clearly when first creating a note this is the only option, a new note is, well...new, and thus all other notes to which you might choose to link are older and thus you are pointing to previous work.

Suppose at time 1 you link 123 -> 121

Fast forward into your future: what happens if you are editing 121 , and find the urge to reference something in 123, do you let yourself do this? You will now have two links (bidirectional) 123 <-> 121.

Does this contradict the ZK principle? If you don't let this happen, I can imagine there being a lot of link maintenance that must go on to prevent yourself from accidentally creating a link which is already a back-link into a regular link?

Thoughts appreciated.

Bradford

Comments

  • edited August 2021

    I let linking happen with zettel with higher UUIDs.

    Only where they call out for creation. Actually, when linking an older/historical zettel to a newer one, like during refactoring (editing), I don't even consider looking at the UUID when nominating a zettel for linkage. It is not in my mental list of criteria. The creation date is not relevant; the ideas are.

    I take each link creation on its own terms. In your example, I would link 123 <- 121 even if there was no link in the opposite direction. Sometimes I end up with a note linked in both directions. Usually not. Linking happens when there are explicit reasons for a link. I don't force it or make some operational rule out of this.

    My rule of thumb is to link zettel that have identifiable morphism.

    By the way, thanks for the pointer to Spivak's olog work.

    Will Simpson
    I must keep doing my best even though I'm a failure. My peak cognition is behind me. One day soon I will read my last book, write my last note, eat my last meal, and kiss my sweetie for the last time.
    kestrelcreek.com

  • @bradfordfournier

    I follow the practice described by @Will as well. In fact, to my understanding, it is essential for the health of your ZK that you do so. But it does require you to occasionally peruse your ZK to consider linkages. One of the times that happens is right after you have created a zettel, in which case you can link to "older" zettels, but another time is when you are reading or reviewing an older zettel and it occurs to you that it makes sense to link to newer ones.

    As an example of the latter, I use the tag "#unlinked", which stays in place on a zettel until I've established at least 2 links to other zettels. Once in a while I review all the zettels with that tag and it may be that a connection to a more recent zettel occurs to me. Why wouldn't I make that link?

    Most of the time, I have no problem thinking of a few links for a new zettel, but sometimes I cannot, which is why I use the "#unlinked" tag. Other times, I'm in the zone writing a series of zettels and don't want to get side-tracked by creating links. So every zettel in that series gets an "#unlinked" tag. I can't always control how long it takes to create those "missing" links. It may be that a week or a month has gone by and there are a lot of newer zettels to consider for linking.

    This idea of "newer" and "older" zettels is a red herring, in a way. If our UUID did not contain a date / time stamp, but was based on some other random system that didn't imply order of creation, would we care?

  • I'm curious, do those of you who are mindful with your notes and links make sure to only link to "previous" notes (those with a lower ID) when creating a link.

    Time is no factor to be considered. Zettelkastens are platonic Miniverses.

    I am a Zettler

  • edited August 2021

    this discussion fails to explain why we always add to older notes, so there isn't really something to reason about.

    When adding a reference directly inside a note you are altering its context. There is a risk involved in this, as it can make an existing connection invalid due to this change.

    I have never heard of this rule. The implications would be that many important connections need to be formed indirectly, as you've pointed out. For example, you need to create a separate note C to express a bidirectional connection between A and B. The advantage to this is that A and B remain unchanged, meaning that A, B and C are self-contained.

    does this contradict the ZKM principle?

    all ZKM principles do not follow this rule
    all ZKM principles can be implemented following this rule


    there are several reasons on how one note connect to another

    • we always connect to existing notes when using Folgezettel

      • violating this rule results in error
    • connections from the broader context to the narrow context are more stable

      • a narrow context has a weaker connection to a broader context than vv.
      • build up knowledge top down
    • connections with similar levels of abstraction are equally stable

      • connections in any direction do not affect stability
    • connections from a narrow context to a broader context are stronger

      • connection from broad to narrow is not always the best choice
      • connections can be equally stable, or more in the other direction
        • build up knowledge bottom up (facts, baby!)
        • we cannot be a generalist forever, at some point we must hold a position

    Clearly when first creating a note this is the only option

    create a new note with no connections.

    I can imagine there being a lot of link maintenance that must go on to prevent yourself from accidentally creating a link which is already a back-link into a regular link

    you don't get it right the first time. Maintenance is unavoidable, though you can learn to reduce the amount of maintenance needed.

    Post edited by zk_1000 on

    my first Zettel uid: 202008120915

  • @bradfordfournier said:
    Does this contradict the ZK principle?

    Is there any such thing as "the ZK principle"? I've never thought there was. I just saw a method for being able to reference and find notes that was quite flexible and could be adapted to what I wanted to do. If something is useful to me, I use it. I'm disinclined to follow any "rules" that get in my way. I see a notes archive as being organic, rather than being a piece of engineering that has to be constructed properly in order for it to work. It ought to be able to grow as it needs to, not in obedience to some abstract notion of what it ought to be. Just my personal view, of course. :)

  • If age of zettel means anything for the links between them, then, well, the information is there. Presumably you have the creation date for each zettel somewhere.

    If a link to a newer zettel would be "wrong" you can always keep that in mind when following such a link. So, if somehow at some time a great revelation occurs to us, how time is now meaningful there, nothing is lost. You just end up with a classification scheme for links.

  • @MartinBB said:

    @bradfordfournier said:
    Does this contradict the ZK principle?

    Is there any such thing as "the ZK principle"? I've never thought there was. I just saw a method for being able to reference and find notes that was quite flexible and could be adapted to what I wanted to do. If something is useful to me, I use it. I'm disinclined to follow any "rules" that get in my way. I see a notes archive as being organic, rather than being a piece of engineering that has to be constructed properly in order for it to work. It ought to be able to grow as it needs to, not in obedience to some abstract notion of what it ought to be. Just my personal view, of course. :)

    Haha! @MartinBB You are a person after my own mind, but then I think we are of similar age and stage as well, which might have something to do with it. An obsession with rules on how to do something is akin to an obsession about managing our time. Both (rules and time management) have a place and provide benefit when used judiciously; they both also can be ignored, at times, in favour of living our lives in a meaningful way (and that of course is another discussion).

  • @GeoEng51 Maybe it is some strange sort of Canadian connection ;) I have distant relatives in BC and Saskatoon. But anyhow, I am glad that I am not alone in some of my considerations :)

  • edited August 2021

    I treat the Zettel ID as immutable and unique, with no additional structure, such as an ordering. There is no requirement that the digraph has at most one link between any two Zettels.

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • edited August 2021

    Example: suppose you want to create a sequence of notes in order. There is no rule insisting that the next note in the sequence must link to the previous note. Why can't the current note link to the next, higher numbered note in order? Notes can be edited after the fact (unless you're building a cumulative hierarchy or you need the ZK to induce the structure of a finite free category for some reason). To follow the rule that all notes can only point lower numbered notes (shades of Russellian type theory again!), one would have to write the notes in reverse order. But by all means, I could be oblivious to endless cognitive vistas with my approach. :smiley:

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

Sign In or Register to comment.