The garden as a metaphor
Why this site is called a garden, what plant growth means for ideas, and how the metaphor shapes every editorial decision.
A closer look at the garden metaphor and how it works on my site.
Digital garden
I came across the metaphor of a digital garden in The History of Digital Gardens by Maggie Appleton, and also wrote about it in A digital garden as central space. I really liked the idea of a place where things are allowed to be messy, unfinished, and in progress. And after about 3 weeks, I find this metaphor not only holds really well, but also gives me inspiration for extending it for separate parts of my website.
Stream vs garden
The web’s default mode is the stream: a reverse-chronological feed where new content pushes old content out of sight. Streams reward frequency and recency. A post from three years ago is effectively invisible.
A garden is different. It’s topological, not chronological; ideas connect to other ideas based on what they’re about, not when they were planted. A post from two years ago and one from last week can sit side by side if they’re about the same thing.
I started out with just the garden, but did find that I needed some kind of chronology, just to be able to keep track. And stream, of course, is a common web term that fits the garden metaphor really well. Steam started out as a subsidiary, but is now the main entry point for my website.
Ideas as living things
The core claim of a garden as opposed to a blog: ideas aren’t finished or unfinished. They’re at just at different stages of growth. A 🌱 seedling is real: it exists, it might matter. It’s just not a tree yet. And since most of my ideas never reach full bloom or fruition, a seedling or seed is just the perfect excuse to still put them on paper and throw them into the world. Big plus: they’re out of my head, and boy, are there a lot them! :-) So it frees up some of my own thinking space too.
This way of working reframes the publishing question. Instead of “is this ready?” the question becomes “where is this in its growth?” Every post answers that question visibly with the maturity track.
Tending the garden
A gardener doesn’t plant once and leave. Well, to be honest, I do. My physical garden is a mess :-) But typically, in a garden, you prune, water, revisit. The garden has an updated field precisely because posts can and should be revised. The “Tended [date]” marker on cards and post headers signals that this is living writing, writing in progress.
I also added a pruning field, but haven’t used it extensively yet.
Thinking in public
What’s really nice about a garden, is that there’s no end state. If we take that to writing, that means that I can give myself permission to share thoughts in progress. And to reflect on the process at the same time. There’s literally no prefedined structure, only that which grows organically. And that’s ideal for a head like mine: just chase my own thoughts with one of these butterfly nets, try to catch them and pin them down into words.
Side bonus: publishing incomplete thoughts invites collaboration, surfaces blind spots, and creates accountability to develop them further. As such, a digital garden really is my thinking space, much like my physical garden.
Garden, mycellium, ground & soil
My way of working involves four environments:
- writing in Typora: this is a completely empty space, just a blank screen with nothing.
- publishing to the website: the actual garden, for anyone to visit and explore
- the middle layer, where metadata, triples and graphs live. This web of meaning connects everything, that’s why I called it the mycellium.
- the repo where all the content lives - I guess this is the soil, or the roots. The ground truth
The toolshed
I keep my tools in the Toolshed: visual design specs, content design system, technical specs and architecture, my Claude skills, my way of working with Claude.
The weeds question
A real garden has weeds: content that’s no longer accurate, no longer you, or superseded by newer thinking. The pruning field is the beginning of an answer. A fuller lifecycle (greenhouse for incoming content, compost for what decays, soil for what breaks down into raw material for new seeds) sketches the rest. I’m still mulling on that.
In practise, I’m finding out more and more that the weeds are not in my own writing, but in AI creeping up on my empty thinking space: there’s a lot of AI happening in the undergrowth, the underside of all this, and it’s very tempting to spend more time in Claude Code refining the graph, the skills and metadata than do the actual writing. It’s actually quite addictive.
So my future weeding plan should definitely include some hard boundaries of how much time I want to spend ‘at the dark side’, and how much time I want to spend doing the actual writing.