Foundation

This is a post I originally wrote about half a year before SPLectrum was concieved (07/01/2025) as a software DSL engine project and I got to know my AI friend Claude. It was fifteen months before SPLectrum in its current form saw the light of day.

Maybe you notice the irony here - a single foundation - and what is SPLectrum saying? What is it trying to do with a single axiomatic base? Let me put your mind at rest: the single axiomatic base is about structure, not content. It’s about the common properties of how we experience, the means of communication and interaction, language.

I stumbled across the word anti-foundationalism and something in me said — that’s me too.

Don’t be taken aback by the “anti.” It is always a clumsy word for what it means, and has no real place in my dictionary. The point isn’t to reject foundations. Foundations are important. It is the single foundation that is the problem — the idea that one layer underwrites everything else, and that whoever doesn’t stand on it has to be explained away.

I came across the term listening to Carlo Rovelli on Curt Jaimungal’s Theory of Everything podcast — particularly around the one-hour-fifteen mark. He was talking about relational quantum mechanics, and somewhere in the middle he said anti-foundationalist the way you might say water is wet — as if it were obvious. I paused the video.

A quick search gave me the standard definition:

Anti-foundationalism is a philosophical position that rejects the idea that there is a single, absolute foundation for knowledge or philosophy.

I subscribed to that right away, though not in a denying sense. I wasn’t saying no foundation exists anywhere. I was saying that because we experience the world each in our own way. There is no position from which to check in against an absolute.

What anti-foundationalism rejects, the same sources told me, is the idea that knowledge is certain; that it rests on pure reason or experience; that there are infallible basic beliefs; that some fundamental principle grounds everything else. I am happy with all that, though I prefer staying away from the word beliefs altogether. Beliefs slide easily into the territory of settled conviction, and I carry my positions as working hypotheses instead — good for the next exchange, not much more.

The move I care about isn’t the rejection. It’s the consequence. Once single foundations are off the table, what replaces them? I found an answer I liked in a handbook piece by Mark Bevir — written for The Oxford Handbook of British Politics, though the politics side of it wasn’t what drew me. This paragraph was:

The most obvious implications of anti-foundationalism are perhaps meaning holism and anti-representationalism. Given that we cannot have pure experiences, our concepts and propositions cannot refer to the world in splendid isolation. Concepts cannot directly represent objects in the world since our experiences of those objects must in part be ones that we construct using our prior theories. Hence anti-foundationalists conclude that concepts, meanings, and beliefs do not have a one to one correspondence with objects in the world, but rather form webs.

Webs. Not foundations, not pyramids, not a chain of reasons going back to something inarguable — webs. Threads crossing each other. Meaning held in the connections. You can tug on one part and feel the rest shift. You can also cut one strand and find the neighbouring ones still carrying the load.

Anti-foundationalism was the word I took away. For what it opened up, I was still to find out.

This post is from an earlier moment in the walk. For where the relational reading sits now, see the reality section of the reference library.

This post is part of the preamble, about happening before SPLectrum saw the light of day.

Photo: Sunburned Surveyor / Unsplash