Interconnections
Thursday March 3, 2022 — Taipei, Taiwan
I recently found that Maya linked to my notebook a few months ago. She wrote:
The vibe is sort of like reading beautiful little booklets, which is wonderful and non-distracting but also not very hypertexty. Their pieces don’t link among each other a ton so far as I’ve read. I wonder if it’s an intentional choice?
This is something I’ve thought about a fair amount. The idea of interlinking one’s own posts has become somewhat popular recently, I think largely spearheaded by Maggie Appleton’s ideas about digital gardening, and driven by the rise of notetaking-turned-publishing tools that have bidirectional linking built in.For some context, I currently use Obsidian to take notes, but I completely ignore the backlinking features, which I’ve found to essentially be gimmicks. I’ve written a bit more about my ideas relating to notetaking systems here. This seems to work for a lot of people,For instance, Winnie Lim, who I’ve been corresponding with via blog, seems to have put a lot of thought into this sort of linking. but I’m not convinced that it’s the ideal thing for me, or in general.
A really core part of how I think about the world is looking for connections between seemingly disparate ideas, which at first glance, seems like a mode of thinking ideally suited to backlinking and building a interconnected web of thoughts. The problem, though, is that a large part of why I find this mode of thinking useful is that you can create useful insights by connecting nearly anything to anything else: the world is fundamentally so highly interconnected that trying to explicitly capture connections between ideas is bound to fall short. Everything is connected to everything else.
Just because something is bound to fall short, though, that doesn’t mean it’s not useful. A more fundamental reason that I’m sceptical about the hype around interlinking is the sorts of behaviour it seems to encourage: structuring your thought around links between existing ideas means that there’s a pull to categorize every new thought as related to some grouping of existing thoughts. This lends a sort of stickiness to categorizations that I’m uncomfortable with — it pulls ideas together that could potentially grow better with more space to flourish on their own.
Another common theme in the “digital gardening”/interconnected notetaking hype is the limitations of temporal linearity. I recently reread How the Blog Broke the Web, which blames the downfall of the web many people are nostalgic for on this linearity. I’m sympathetic to this, but I’m not sure that I agree, on the whole. While brains may work nonlinearly, time offers no such affordances: we only ever have our past thoughts and experiences to draw on, never our future. It’s tempting to think that breaking this externally enforced temporal linearity will make one’s ideas more legible, but I’m not convinced: I find it easier to understand writing where the influences are clearly and simply laid out. If I change my opinion on something, it’s much clearer to simply write what my new opinion is, than to try to edit and untangle everything I’ve written in the past. As a reader, that’s easy to understand: we all know that opinions can change over time, and people can get a clear view of what I believe now, by reading my recent writing, as how I came to those beliefs, by reading through my past posts. It may not be beautiful, but to me, it feels honest.
So far, this would suggest making a tree of links growing into the past. I think that’s generally a good modality, but even that often doesn’t feel quite right to me. The question of whether I link to one of my old posts or not isn’t answered by whether they’re related, but by whether the act of linking serves a purpose. If someone follows the link, will they come away with a better understanding than without the link? I find it’s often better to summarize a few ideas, rather than link to the place I first explored those ideas. That’s for a few reasons:
- Often, my thinking on the topic will have evolved, and the original post no longer accurately reflects my ideas.
- Sometimes, the original post is written in a very specific context, and I want to talk about the idea in a different context, where thinking about it in the first context will just be distracting.
- If reading links is required to understand the post, and there is heavy interlinking, that means that reading the entire site is required to understand any of the ideas. Expecting people to read links is putting a undue burden on them, and if I’m not expecting people to read the links, I’d rather allow them to explore my writing under their own direction than push them onto specific paths. Maybe that way, they’ll find connections I wouldn’t even have thought of!
On the whole, I feel that this sort of interconnection is good for some types of things, but not for others. For me, it feels good when I’m creating an artifact that changes slowly, describes something I understand well, and will stand on its own: documentation, for instance, benefits heavily from this sort of interconnection. Implementing this interconnection well enough to be useful imposes costs, though: it can make it feel more difficult to write about topics that aren’t connected to the superstructure of ideas I already have, and which don’t slot neatly into existing taxonomies.
I don’t mean to be too down on interlinking: it clearly works for a lot of people. I just find that I’m generally happier writing linearly, and trusting that the important connections will reveal themselves serendipitously.