< More Posts
Published 16 May 2021 at Yours, Kewbish. 2185 words. Subscribe via RSS.
I can’t seem to find it again, but earlier this month, I saw a video on Twitter of a proof-of-concept zooming essay model. It featured a straightforward explanation of some maths, but what really struck me was the layering of the concepts: each ‘sentence’ or major phrase was underlined in a different colour, and when clicked, would expand into a longer, more intuitive layman’s explanation of the topic. Not every word was expandable, but, for example, a sentence involving a couple formula definitions would lengthen into several more phrases developing the background knowledge. That ‘second level’ explanation in turn contained a little bit of theory based on some other assumptions, so the reader could click again to expand key points in that explanation, and so on. I think this continued for three or four levels, and what I found most interesting was that this augmentation wasn’t limited to individual phrases. If I recall correctly, one of the sentences involved the definition of some probability statistics,which was also touched on later in the paragraph. Clicking once on the first definition would expand topics later on in the writing, so the entire piece of writing had several levels of explanation ‘advancement’, going from expert complexity, with lots of notation and terms; all the way to something comprehensible by even those with no maths background.
Imagine the possibilities if these were extended to mainstream explanations and articles. With most blog posts (including mine, newsletters, and platforms like buttondown.email, this level of expansion and ‘for later investigation by the reader’ is already possible with footnotes, but not quite to the level that was demonstrated in the GIF. Sure, you can link relevant resources and direct the reader to other related ideas, but with the Twitter example mentioned above, there’s an extra dimension of presentation control. The author can fine-tune the level of jargon and generalizations present in each expansion, and has more power to present their essay or text as they’d like it to be experienced.
If you read any Wikipedia article, you’ll find that the first paragraph or so is a simple, clear summary of the topic at hand. Further links lead the reader to additional materials to increase their understanding of the depth of the topic, but skimming that first chunk is enough to get a gist of the topic itself. I’m reminded of the idea of incremental learning1, where instead of starting from complex theory step by step, you build a general understanding of the topic first. You might have to make a couple overall generalizations and assumptions, but the reader’s left with a basic grasp of most of the topic. Then, each iteration of a text can build on the level before, clearing up a specific point or introducing new vocabulary. At each step in the process, the learner has a working understanding - it might not be the deepest or most accurate, but it’s there. This works well with a sort of reversed version of the Twitter example (I’ll just call it that since I forgot the actual name of the creator and can’t find the video anymore) - using each expansion to further comprehension, not explain a point. Same concept of levels of knowledge, different structures.
This post is an exploration of my thoughts on theories that expand beyond just lists and dry bullet points, how to efficiently pursue idea connection, and how I approach the writing process. I really like the idea of infinite subexplorations and the ‘zooming’ magnification present in that demo, and I don’t entirely know why. Perhaps it’s that it affords the reader more opportunities to customize the text to their own reading and comprehension levels, or maybe it’s that everything is structured very logically, with an overall outline that can be (almost) infinitely expanded. Honestly, thinking back to the long digressions I had on click-holes, I think it’s just what appeals to my brain the most.
This type of thinking ties basically the whole non-linear notetaking workflow and recent networked thought revolution together. Processing information in a way that doesn’t involve immediately writing down final notes or content sequentially, or what I call non-linear thinking, is something that’s been gaining traction recently. People praise it for mimicking how their brain works, and are surprised with how much less restrictive the entire system is. Thinking like this is inherent in models like the Twitter example, where each ‘layer’ of text interacts with others but stays independent (for example, you can expand one set of definitions, while leaving another about a topic you’re more comfortable with as-is).
There are many different manifestations of platforms that I think fall somewhere under the non-linear umbrella, such as infinitely nesting notes, linked systems like the ever-popular Zettelkasten or Roam Research graphs, and outlining tools. Tools available in this area might overlap in one or more of the areas above, but the main point is that things don’t (have) to happen sequentially or from start to finish in one go. Thoughts can have intermediate layers, and iterations can stand alone while being incomplete and marked for further revision. Systems like this work to reorganize and restructure large amounts of information quickly and efficiently. Think back to trying to edit context-heavy sentences that draw on its neighbours to provide transitions and such, versus copy-pasting a bunch of bullet points around, while retaining any links those may have to other points. Off the top of my head, I can think of two main players in the space.
While I’ve never used Workflowy as a daily notetaking driver, I’ve heard it described as one of the original implementations of non-linear thinking. One of its biggest selling points is the infinite nesting and pagination present in its model. Each point, termed a ‘bullet’, is an entire page - sort of a lower level of what’s available in Notion and Roam. Each bullet can contain infinitely many other bullets, which opens up the possibilities for a lot of outlining-style and level-zooming type interactions. On its marketing page, it states that:
No other document has an infinitely deep structure that lets you choose the exact focus level you want. which directly echoes the whole layers concept that I think is an interesting platform to work with.
Roam Research is slightly different, but has equal following (especially on Twitter, with the whole #roamcult shebang) and vaguely similar aims. Also one of the most popular tools for building networked knowledge bases, its main selling point is its whole graphs ecosystem and the blocks model that it so popularly uses. It’s a cult favourite in the Zettelkasten community, and has been lauded for completely revitalizing how people work. I’ve also never personally used Roam, so while I can’t speak directly to its apparent life-changing potential, I think it marks something interesting: the shift in awareness and prevalence of non-linear thinking tools.
Over the past couple months at school, I’ve realized that I keep returning to one ‘way’ of doing schoolwork. Every time I create a new document for some homework assignment and finish all the setup of 12pt Times New Roman and double spacing, I start by copying over the guiding questions2, which I find is a nice way of sparking thoughts and feeling a lot more productive than I really am. Oh well. Then, I start going through each question - and where I used to start by going systematically in with fully-fledged answers right away, I tend to just create a new bulleted list and come up with my main points and subpoints. I call this system outlining, and it’s been pretty surprising how much it’s taken over how I work.
Outlining is a bit of a one-dimensional version of the general ecosystem of networked and infinitely capable knowledge management tools. It dilutes (or well, not dilute, since the concentration of thought is the same, if not more) the complex process of creating full notes with perfect grammar and fully thought-out ideas into a more iterative process. I can quickly lay out each of my points, while not having to worry about grammar, coherence, or even having a proper idea. I tend to use super casual language, and just go ahead with stream-of-consciousness style writing, full of ‘like’s and ‘whatever’s and such. I don’t have to worry about crafting each sentence until I’ve actually developed my idea and figured out what I actually want to say, and it’s a lot easier to work through several levels of iteration in the process. At this stage, things can still be reordered and ideas can be more developed very efficiently, and you’re not locked in by any concrete ‘work’. Once I’m done with my outline, I go back to restructure the text into full sentences and paragraphs, taking out all the slang, replacing repeated words, and refining the overall thoughts.
This kind of ties back to ‘conversational notetaking’: I find that with just typing out whatever I feel, I end up writing vaguely sentence-like thoughts, but they’re still less focused on the wording and more on the idea. Feedback loops are very short, and I can work to develop my ideas more quickly than if I’d have to go directly in with worrying about grammar and stressing over the correct language to use. Oddly, I’ve found that I end up using most of the right words anyways, but I’ve found that the writing process still goes by much more quickly than if I’d started with an empty document and started grinding through full sentences. At each iteration of the outlining process, I have a full document - all my ideas at each point in time are laid out on the screen. An advantage of this method over the more linear ways of working is exactly that - the layering and successive development makes each step more efficient, and it definitely helps with avoiding that helpless feeling, since each thought is at a technically completed state with each stage of the system.
Strangely enough, even having found that my brain really likes the intermediary structures of outlining and the networked capabilities of a linked notes database, I still write blog posts the ‘normal’ way: just start to end, composing full sentences as I go. I guess part of it is that articles here feel a lot more casual and stream-of-lightly-edited-and-filtered-consciousness here as opposed to more formal schoolwork, but I think it’s also that I have a pretty good understanding of what I want to write about every time I sit down to work through a post. I don’t have to struggle to structure my thoughts more, so while I lay out a couple of markdown headers with work-in-progress headers and put down a couple key bullet points, I largely just go ahead without a formal ‘outlining’ process. I do jump around a bunch when sentences pop up to the top of my head, but I really have no idea why I haven’t tried outlining a post the way I’d write other work. Something to consider for the next post.
Speaking of posts, it’s been a while, and it might continue to be a while. It feels weird to be apologizing (?) with each post, but I genuinely don’t feel like I’ve been putting in enough time to write, even when I do have ideas. School was expected to start winding down roundabout now, but it’s started ramping up with final projects and such instead, and I’m looking ahead to prepare as best I can for UBC. The next couple weeks look pretty packed, but I think I won’t bother writing unless I make it a personal assignment, so I will certainly attempt to slot in more timeblocks in my calendar. This blog is a place I feel that I can sort of finally expand on and formulate my own thoughts, and it serves as a reminder of what I was thinking of each week in time - I’d like to keep it up as much as possible. Oh well - next week we’ll see what’s going on, and we’ll see if I come up with another couple coherent thoughts or not.
I guess this is why the education system (at least in Canada) is structured in such a spiralling method, where we cover a topic, then go up a grade and do essentially the same thing, but in an ever so slightly more advanced way. I’m not entirely sure why learning in layers (such as a textbook organized in the Twitter demo-esque way) is appealing while I find the spiral method awfully slow. I suspect it might have something to do with tighter feedback loops and having the ability to control your own speed in one (self-learning) but not the other (school). An interesting thought. ↩︎
Generally, most of my work ends up being stuff like reflections, essays, question / problem sets, which apply well to my system. With things that are more free-form, this process still works well, but there’ll be an intermediate step of brainstorming actual creative directions and ideas instead of just copy pasting any requirements over. ↩︎
- Yours, Kewbish
< More Posts