Advaita Vedanta approaches truth by removing, not adding. It doesn’t build up a description of the self piece by piece until you arrive at a full picture. It works the other way — neti neti, “not this, not this” — stripping away everything that isn’t it, on the assumption that what’s left when you can’t strip anymore is closer to real than any list you could assemble.
I’m not a scholar of this, and I don’t want to dress up design work in robes it hasn’t earned. But that one move — arriving at something by removing rather than adding — is the most accurate description I have for how the good decisions on this site actually got made. And unlike most philosophy, it turns into a concrete practice.
The additive instinct is the default, and it’s usually wrong
Almost everyone’s reflex, mine included, is to treat “I’m not sure if I need this” as a reason to keep something. Keep the extra button, just in case. Keep the second call-to-action, just in case. Keep the ambient animation, because empty feels like a bug.
The subtractive method flips the burden of proof. The thing doesn’t get to stay because removing it feels risky. It has to prove it’s load-bearing. And the fastest way to find out is to ask the removing question instead of the keeping one:
| The additive question (keeps things) | The subtractive question (cuts things) |
|---|---|
| “Could this be useful to someone?” | “If I delete this, what actually breaks?” |
| “Does it look a bit empty without it?” | “Is the emptiness a problem, or just unfamiliar?” |
| “What else could we add here?” | “What’s the least this could be and still work?” |
| “Does it hurt to leave it in?” | “Am I keeping it because it earns its place, or because cutting feels risky?” |
The column on the left never runs out of answers — you can always justify keeping something. The column on the right terminates. It gives you an actual verdict.
A worked example
The homepage’s background field renders nothing when no one’s near it. That wasn’t the first version. The first had a slow, constant ambient drift, because stillness felt like something had failed to load. It took a few rounds of asking “what actually breaks if I remove this motion” to land on the honest answer: nothing breaks. The field is better as stillness with movement available inside it than as movement that never stops. (I wrote up the mechanics of that in Building a Cursor-Reactive Particle Field.)
That’s neti neti in miniature. The field isn’t the particles, the same way the method isn’t the list of things being negated. It’s what’s underneath, once you stop adding to it.
I’m not claiming your interface is enlightened
To be clear: I don’t think an interface is spiritual, and I’d be suspicious of anyone who did. The borrowed idea is narrower and more useful than that. It’s that honesty usually lives on the subtractive side — that most things you’re unsure about are things that don’t need to be there, and the discomfort of removing them is not the same as the removal being wrong.
The good news, every single time I’ve actually done it: almost nothing true was lost. That part is worth sitting with, because it’s the part the additive instinct never believes until it sees it.
A subtraction practice
- Open your current project. Find one element you've been unsure about. Instead of asking whether to keep it, ask what breaks if it's gone.
- Actually delete it — temporarily. Look at the result for a full minute before judging. First reactions to emptiness are almost always "wrong," then they settle.
- If nothing broke and nothing true was lost, leave it deleted. Notice the specific discomfort you feel doing this — that discomfort is the additive instinct, not evidence.
- Repeat until removing the next thing genuinely does break something. That thing is load-bearing. Now you know your real structure.
- Next time you're tempted to add something "just in case," write down what would break without it first. If you can't name anything, you have your answer.