Zettelkasten Forum


Using different kinds of tags

I started maintaining a zettelkasten-like system last week. However, I haven't really bought into all of Lehmann's 'rules' and processes, nor all of the most common conventions that seem to be canonical features of digital implementations of zk systems. One of my deviations pertains to tagging. I'm writing this post to share my own processes, to solicit feedback, and to help make this more robust. Writing things out for others generally helps me work out my thoughts on messy or systematic problems.

So basically, I find value in maintaining different sets of tags. I wouldn't go as far as calling them categories, more like classes or kinds. So far I only have two classes, and I suspect that I will need a third kind, but I'm not really sure how I will go about doing that yet.

It might also be beneficial to describe the context of my work and the general purpose of this system before going on further. I wrote another post earlier on, but I don't think it really does justice to my goals and processes. I'm doing a qualitative data analysis project for my dissertation, and I'm currently reviewing the data while creating systematic memos that help me familiarize myself with the material, determine areas to focus on, and potentially inspire new ways to approach the data I've collected. I'm not really using ZK to review published work, but my use case entails a similar leap from the perspective of author/subject to my own understanding of the phenomena of interest. I'm identifying processes, activities, attitudes and figurations as elicted or enacted by subjects, and fitting those into my own theoretical perspective (which are first loosely organized as sensitizing concepts, and then grounded as more definitive or refined concepts through qualitative coding).

I use # to flag abstract terms, which currently correspond with 'sensitizing concepts'. According to Kathy Charmaz:1

A sensitizing concept is a broad term without definitive characteristics; it sparks your thinking about a topic. Sensitizing concepts give researchers initial but tentative ideas to pursue and questions to raise about their topics. ... Guiding interests, sensitizing concepts, and disciplinary perspectives often provide us with points of departure for developing, rather than limiting our ideas. Then we devise specific concets through studying the data and examining our ideas through successive levels of analysis.

I typically include # tags in a tags section at the end of each note, and rarely within the meat of the memo itself.

The other kind of tag I use is $, which prefixes significant discrete entities or actors that the subject either mentioned or referred to implicitly. These tags are effectively used to index the entities that authors deem significant to the main ideas or problems that they elicit. I try to use the author's own words when using $ tags as often as I can, though I'm also trying to maintain my own taxonomy of terms that I could potentially relate to the author's own words if I find that they do not adequately capture the thing I identify as relevant. Although I haven't actually made any taxonomic links yet, I could potentially accomplish this either by applying a type of link (more on this later, maybe in another post) to a note where my own term is defined, e.g.: {.link:taxonomy}[[12345678]], or by referring to it informally or quasi-formally by identifying and/or locating the specific document or context where I define the term with as precise provenance as I can.

$ tags are mostly applied within the meat of my memos, but appear in the tags section sometimes too.

I'm also on the fence about another kind of tag, which I currently tend to lump with # tags. They are longer phrases, with words separated by hyphens, which kind of bridge the grounded and abstract perspectives. Here are a few examples:

#Establishing-collaborative-boundaries-and-roles
#responsibilities-of-director
#maintaining-connections-with-the-material

These are ideas that I want to hold onto as I go, which appear from time to time in multiple memos or instances of multiple interviews with various people. They tend to represent relationships between entities or actors. Maybe I can find a way to leverage the connecting words ('of', 'with', 'and', etc) in the tagging system somehow. I'm having a hard time keeping these tags consistently organized in a non-redundant way, while also accounting for subtle variations that may alter the meaning or relevance of different uses of these tags. They seem to be very context-dependent. One term I like using to describe them is 'topics', though I'm not totally committed to it yet.

I wonder what others think about all of this.


  1. From the second edition of Constructing Grounded Theory (2014). Although I do not adhere to all of grounded theory's primary tenets and practices, I selectively draw from several grounded theory principles, similarly to how I pluck the parts of ZK that work well for me. ↩︎

Comments

  • Very interesting indeed.

    A sample might be helpful.

    @mtl_zack said:
    ... reviewing the data while creating systematic memos that help me familiarize myself with the material, determine areas to focus on, and potentially inspire new ways to approach the data I've collected ... my use case entails a similar leap from the perspective of author/subject to my own understanding of the phenomena of interest. I'm identifying processes, activities, attitudes, and figurations as elicited or enacted by subjects, and fitting those into my own theoretical perspective (which are first loosely organized as sensitizing concepts, and then grounded as more definitive or refined concepts through qualitative coding).

    This sounds so much like an ideal Zettelkasten workflow. This is a goal to shoot for and if successful, wonders of wonders. This is well said and formulated.

    I use # to flag abstract terms, which currently correspond with 'sensitizing concepts'. ... I typically include # tags in a tags section at the end of each note, and rarely within the meat of the memo itself.

    Two classes of tags #(abstract terms) and $(significant discrete entities or actors). This is a style of tagging I hadn't considered. Is your "zettelkasten-like system" focused on only one project or does it have the potential for greater integration into your knowledge work? I use the term knowledge work loosely.

    Although I haven't actually made any taxonomic links yet, I could potentially accomplish this either by applying a type of link (more on this later, maybe in another post) to a note where my own term is defined, e.g.: {.link:taxonomy}[[12345678]], or by referring to it informally or quasi-formally by identifying and/or locating the specific document or context where I define the term with as precise provenance as I can.

    This is confusing to me, I don't understand your meaning of taxonomic links. Tutor me, please.

    tags are mostly applied within the meat of my memos, but appear in the tags section sometimes too.

    I find interstitial tags to be most helpful. First I can create the tags on the fly. Second, I don't have to repeat the words again. Third, I know by the tag being interstitial in the memo, exactly where and what the tag references.

    Here is my example - probably not your cup of tea but it displays the idea. I'm still a work in progress and don't always spend the time I need to flush this out.

    I'm also on the fence about another kind of tag, which I currently tend to lump with # tags. They are longer phrases, with words separated by hyphens, which kind of bridge the grounded and abstract perspectives. Here are a few examples:

    #Establishing-collaborative-boundaries-and-roles
    #responsibilities-of-director
    #maintaining-connections-with-the-material

    I too am experimenting with longer descriptive tags try to find some use case for tagging besides project tagging. To be honest, I find tagging mostly tedious, outside of project tagging (see recent forum thread). I want to think about tagging from the perspective of what use they'll be in the future. So far, most tags are proving to have no future utility. YMMV.

    I wonder what others think about all of this.

    These are my humble thoughts.

    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

  • Thanks for your comments @Will! This was initially meant as a response to your recent post in that other thread, but then after writing a few lines I deciced to make it its own thing.

    @Will said:
    Is your "zettelkasten-like system" focused on only one project or does it have the potential for greater integration into your knowledge work?

    I think I will keep it focused on this project, but since this is my first zettelkasten it's hard to tell how things will go in the future. If I do it on a project-by-project basis I would probably try and come up with a way to merge finished projects into a unified archive. I imagine that this will probably be a stage of work in itself, involving reviewing prior work and establishing links to older material.

    @Will said:
    This is confusing to me, I don't understand your meaning of taxonomic links. Tutor me, please.

    I have a structured list of terms and definitions that I created with my research questions in mind. The terms are organized hierarchically within a few classes, so definitions are actually technically 'scope notes'. So for example, under a stub for [People], I have [Individuals], [Personal traits], [Roles] and [Collectives]. Each contains a kind of thing that is also within the realm of [People], but are distinct among themselves. This can be drilled down with even more granularity; so under [Collectives] I have [Social groups], [Epistemic groups], [Project teams], [Organizations]; and under [Organizations] are [Museums], [Universities], [State organizations] and [Companies].

    Stubs are the terms in square brackets, and are like containers, not things in themselves. They may include instances of those things. So [Universities] might include University of Toronto, McGill and McMaster, and these would then also be within the scope of [Organizations], [Collectives] and [People]. This is an example from the descriptive component of my taxonomy that will largely complement $ tags, while I will also probably refer to terms under a theoretical wing for some of the # tags.

    I only realized after writing this example that I don't have scope notes for these particular stubs, but here is one for the term [Hardware]: "The machines, wiring, and other physical components of a computer or other electronic system." Note that the term has other meanings for other people and purposes (as in the construction industry), so it's useful to specify what I'm referring to rather than just use a lexical query.

    I haven't quite figured out how to bring this into the zettelkasten system yet. But I should probably be assigning each stub to a note with its own UID, and include the scope note (definition) and a special character in the title to indicate that it is a taxonomic term. I need to find a way to automate the initial creation of these ZK notes, since they currently only exist in a proprietary database (though exportable to excel).

    Specifying a type of link is just an idea I've been playing around with, because I might want to just find links that specifically serve a purpose of indicate that a term is equivalent to a term that I already described in my taxonomy. Or I might want to search for links that serve some other specific purpose. {.link:xyz} just seems like a potentially useful for looking things up, and is inspired by an informal semantic markdown specification I came across: http://blog.sparna.fr/2020/02/20/semantic-markdown/. As I wrote above, I'll probably write something up about that once I actually encounter that kind of application or need more frequently in my work.

    @Will said:
    I find interstitial tags to be most helpful. First I can create the tags on the fly. Second, I don't have to repeat the words again. Third, I know by the tag being interstitial in the memo, exactly where and what the tag references.

    I don't always use the term in the body of the memo. But I think I will try to move them up closer to the parts of the note that they pertain to, rather than in a section at the end.

  • @mtl_zack said:
    I think I will keep it focused on this project, but since this is my first zettelkasten it's hard to tell how things will go in the future. If I do it on a project-by-project basis I would probably try and come up with a way to merge finished projects into a unified archive. I imagine that this will probably be a stage of work in itself, involving reviewing prior work and establishing links to older material.

    I hope you'll let this project grow it infect your whole learning life. I have found most of the value of Zettelkasting is at the intersection of 'projects'. I've learned about how Dog Training, Nutrition, Neuroscience, and Micro Habit Formation all blend together. Something I wouldn't have otherwise.

    @mtl_zack said:
    I have a structured list of terms and definitions that I created with my research questions in mind. The terms are organized hierarchically within a few classes, so definitions are actually technically 'scope notes'. So for example, under a stub for [People], I have [Individuals], [Personal traits], [Roles] and [Collectives]. Each contains a kind of thing that is also within the realm of [People], but are distinct among themselves. ... Specifying a type of link is just an idea I've been playing around with, because I might want to just find links that specifically serve a purpose of indicate that a term is equivalent to a term that I already described in my taxonomy. Or I might want to search for links that serve some other specific purpose. {.link:xyz} just seems like a potentially useful for looking things up, and is inspired by an informal semantic markdown specification I came across: http://blog.sparna.fr/2020/02/20/semantic-markdown/. As I wrote above, I'll probably write something up about that once I actually encounter that kind of application or need more frequently in my work.

    Thanks, this sounds like a top-down approach. Like trying to determine the ends before one starts. I see you quandary about "... maintaining a zettelkasten-like system".

    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

  • @mtl_zack said:

    This type of tag doesn't work in iA Writer;
    #Establishing-collaborative-boundaries-and-roles

    The one that works is using underscores #Establishing_collaborative_boundaries_and_roles

Sign In or Register to comment.