Zettelkasten Forum


Commentary on NIDs: Noun IDs

What could possibly work as a Word IDs (WIDs)?

Nouns really. Any person, place, or thing. (Maybe this subset could be called Noun IDs: NIDs).

  • Things: [[Hormesis]] could be a good candidate for a NID. The note would be my personal wiki entry. Then, if I want to write about Hormesis separately, it would be titled 202005089999 Interesting Hormetic Happenings.
  • People: [[John Doe]]
  • Places: [[Rome]]

One common argument for some NIDs

It works well for readability.

ie [[Antifragility]] is related to [[Hormesis]]. This is clean and readable as opposed to: [[201605089999 Antifragility]] is related to [[201605109999 Hormesis]]

One argument against having NIDs

It strips a note of its Temporal Context (in which I pitched a case for above).

However...

NIDs may not really benefit from having a DTID "Temporal Context" in the filename. (If not in the filename, you should always have the DTID somewhere within the file of course!).

ie [[John Doe]] is John Doe and that's not changing. If you meet with John Doe, then those notes become [[201805089999 Meeting w John Doe]] and [[202002089999 Meeting w John Doe re Frisbee]]

No problem I see. What problems might come from this?

Comments

  • If you use the NIDs a search for a NID would return bloated results. A search [[Hormesis]] for example would return any Zettel that has hormesis in your title and/or content. (depending how you search)

    I am a Zettler

  • I use words (or titles if you may) for IDs for my notes, almost exclusively, and I think it works quite well for my needs.

    I've followed most of the discussion about IDs, but I can't really say that I have seen anything mentioned, where I felt I missed out on something. Remember, the same purpose or need can be fulfilled in many ways.

    IDs must be unique and so must my titles then. I love this! It forces me to put effort into naming and communicating what a note is about. Frequently, I can't figure out a good name and I identify this as a failure of being able to capture the essence of the note. Typically, I'll just add the content to an existing note. When I revisit it later, I might be able to formulate myself more clearly.

    In a pinch, I've simply added a datetime ID to notes if I really needed the uniqueness but couldn't find a good unique name. Yeah, you can mix ID systems to fulfill different purposes.

    I should also say that the software I use (TiddlyWiki) handles things like uniqueness, renaming etc. for me. I'm some will argue that it is a software dependency and that I will be sorry in many years when I'm locked in. I don't think it is so bad.

  • @sfast said:
    If you use the NIDs a search for a NID would return bloated results. A search [[Hormesis]] for example would return any Zettel that has hormesis in your title and/or content. (depending how you search)

    Why is that bad? From my use of my collection of notes, that has resulted in numerous surprising connections and ideas.

  • edited May 9

    One downside is that searching with unique IDs allows for a kind of rudimentary backlinking, i.e. you only get results of directly interlinked notes. Meanwhile, [[hermesis]] would get you a disorganised cluster of any notes that contain that word, even if they weren't directly linked with each other. This connection between notes is semantically weak, but syntactically on the same level as direct linking. In other words, when you link directly with a unique ID, you consciously specify a semantic value to the connection. Using the same word in different notes, on the other hand, might be purely unintentional. Obviously, both kinds of searches can lead to productive associations, but for clarity it might be better to have a way to keep them structurally separate.

  • edited May 9

    One problem that's immediately apparent to me, and that I think weakens the readability benefit, is the existence of homonyms. I do research in philosophy and I can see it would be a problem particularly for that field considering how much philosophers modify and disagree about concepts. But I imagine it would happen in other disciplines.

    For example, pretty much every major metaphysician and ontologist from Aristotle to Hegel has their own concept of substance. In addition there is chemical substance, which is not totally unrelated to the philosophical concept, but definitely distinct. If I were interested in doing an intellectual history of the concept "substance" I might make a "substance" structure note. With NIDs I would link to this simply as [[Substance]]. But when I'm talking about substance in a note about, say, Spinoza, I wouldn't want to link to my massive [[Substance]] note which is cataloguing the entire history of the concept of substance; I'd just want to link to a note on Spinoza's concept of substance. So I could make a separate note called [[Spinoza's substance]], but if I keep doing that I could end up with sentences like:

    [[Baruch Spinoza]] identifies [[Spinoza's substance]] with [[Spinoza's God]] and [[Spinoza's concept of Nature]].

    The date-ID alternative could look like this:

    Spinoza [[202005090853]] identifies substance [[202005090854]] with God [[202005090855]] and Nature [[202005090856]].

    I find the date-ID format less intrusive.

    Something similar could also occur with names. I might end up with notes on Francis Bacon the philosopher and Francis Bacon the painter. I'd have to disambiguate somehow, which would inevitably be something like [[Francis Bacon (philosopher)]] and [[Francis Bacon (painter)]]. With date-IDs I can just have [[202005090857 Francis Bacon (philosopher)]] and [[202005090858 Francis Bacon (painter)]] and be free to use "the painter Francis Bacon," "Francis Bacon," or just "Bacon" in my sentences while still being able to link.

    In my case, that's why I avoid them. Whenever I linked I'd have to use the fully disambiguated concept-term in my sentences. With the amount of homonymous concepts I'd have to disambiguate (e.g. substance, signification, sense, reason, idealism, functionalism, nominalism, etc), it would intrude on my writing too much.

    EDIT: I confused "homonyms" with "synonyms" (:

  • @Taylor said:

    [[Baruch Spinoza]] identifies [[Spinoza's substance]] with [[Spinoza's God]] and [[Spinoza's concept of Nature]].

    The date-ID alternative could look like this:

    Spinoza [[202005090853]] identifies substance [[202005090854]] with God [[202005090855]] and Nature [[202005090856]].

    I find the date-ID format less intrusive.

    This is how I would do this.

    Baruch Spinoza [[202005090853]] identifies Spinoza's substance [[202005090854]] with Spinoza's God [[202005090855]] and Spinoza's concept of Nature [[202005090856]].

    Best of both worlds. This way if I want to change the verbiage referencing a Zettel in this particular zettel I don't have to change the title of the referenced zettel and worry about broken links anywhere. I could write the following and get the same references not worrying about link breakage.

    Baruch Spinoza [[202005090853]] identifies the substance [[202005090854]] of his notion of god [[202005090855]] with the development of his concept of nature [[202005090856]].

    Will Simpson
    I'm a Zettelnant.
    Research: Rationalism, Zen, Non-fiction Creative Writing
    kestrelcreek.com

  • @Will said:

    This is how I would do this.

    Baruch Spinoza [[202005090853]] identifies Spinoza's substance [[202005090854]] with Spinoza's God [[202005090855]] and Spinoza's concept of Nature [[202005090856]].

    I don't think I'd ever need to write a sentence with that much redundancy, but the indifference of date-IDs to the contents of the notes still allow you to do it if you want :smile:

  • One thing I'm trying to improve in my Zettelkasten is having an appropriate context surrounding the reference to another note. I can't do this with just a [[UID]] and there are just too many problems with using a note's title for a UID. So I've settled on using a [[UID]] and verbiage to describe what will be found in the referred to note. This usually is the title when a note is about just one idea and the title expresses that idea but sometimes more clarification is warranted or one of my notes may express more than a single idea. (Don't tell the Zettelkasten gods!)

    Titling notes is a gap that I'm continually working on. I find myself changing note titles more and more as I create references, my domain knowledge improves, and I find better naming strategies.

    Will Simpson
    I'm a Zettelnant.
    Research: Rationalism, Zen, Non-fiction Creative Writing
    kestrelcreek.com

  • @henrikenggaard said:

    @sfast said:
    If you use the NIDs a search for a NID would return bloated results. A search [[Hormesis]] for example would return any Zettel that has hormesis in your title and/or content. (depending how you search)

    Why is that bad? From my use of my collection of notes, that has resulted in numerous surprising connections and ideas.

    You cannot link to that specific note any more. The bigger the Zettelkasten the worse the problem will be. If your search returns you just 5 Zettel there wouldn't be such a problem. But as time passes the search results grow and the specific link becomes less and less specific.

    I am a Zettler

  • edited May 16

    @sfast said:
    You cannot link to that specific note any more. The bigger the Zettelkasten the worse the problem will be. If your search returns you just 5 Zettel there wouldn't be such a problem. But as time passes the search results grow and the specific link becomes less and less specific.

    This is bizarre. Of course, you can link to a specific note. It has the exact name in the link.

    Sure, if you search. Yes, then there might be many notes turning up -- that is a common (and often desired) result of a search.

    But of course you can link to a note if you use the title as the ID. Or else you know more about my tools than I do ;)

  • Oh, yes. You are using tiddly wiki, don't you? Haha. Ok, I have the plaintext-bias. :smile:

    I am a Zettler

Sign In or Register to comment.