Author: Daniel B. Newman
Published: August 18, 2026

I ended the second article by asking how I could build anything without quietly putting the ground back.

I was still following a thought experiment, not treating fRadical Emergence (fRE) or fPerspectivalism (fPism) as true. The question was practical. fPism named the epistemic implications of fRE. I wanted to know what I could build from them.

For me, philosophical prose kept turning into semantic engineering. Every term I used to stabilize a frame opened another question about what that term assumed. I needed an idiom in which those assumptions would change the build.

I was already working with LLMs to explore what I could build from process philosophy rather than using it only to critique reductionist frames. I asked them how they worked and what they made of the ideas I was developing. Working with them gave me a responsive computational medium: I could change the starting assumptions, encounter what followed, and allow their responses to change my questions. Computation, in the form of LLMs, was already participating in the design of the logic I was building. I chose computation itself as the idiom in which to continue, working through the manipulation of symbols, semantic or otherwise.

Once I looked at computation through fPism, I saw how much I had been taking for granted. I had been expecting information to remain information from one operation to the next. I had been treating the transportability of state as foundational. State had to remain stable long enough to be compared or changed. A transformation had to preserve enough identity for something to remain recognizable before and after it occurred. I had been assuming values, places, and runtimes were already there and that a shared reality held them together.

Those expectations had served me well. Within the thought experiment, however, the shared ground they required was ontologically unavailable. I was curious to know what I could build under those conditions.

Even then, I kept trying to find a perspective from which I could reconcile one cut with another. I looked for a common state, language, or algorithm that could hold the cuts together. I tried to make them interoperable through a universal runtime. Each perspective I introduced became another local cut. Any relation it made available was available only within that cut. It could not stand outside the cuts and reconcile them.

Within the thought experiment, declaration became my starting point.

At that point, the f prefix I had used in fRE and fPism became active in the build. I used f-terms to define logics and conditions that inherited semantics could not adequately hold. Their declared semantics changed the questions the LLMs and I could ask and what could become available as we built.

I declared the local coherence through which a build could take form and called it an fField. I was building within the cut while simultaneously building the cut itself, which allowed me to keep it unreconciled and irreducible. The declaration, the bound, the distinctions, the relations, and what could count as a participant or runtime took form together. Local meant situated, not small. Scale did not make a cut nonlocal.

The computational question was now one of intelligibility: what could become available here, under these declared conditions, as something that could participate, relate, or matter? The logic that followed was not what I had expected. I named the resulting computational paradigm Bounded Field Computing (BFC).

Within BFC, fComputation could arise, if at all, with the conditions that made distinction, relation, and relevance locally coherent. Those conditions changed the questions I could ask of computation and, with them, the answers that could become available.

Within fComputation, what computes, what is computed, and what can matter take form together. State, information, relation, transformation, and meaning can become locally efficacious. 

State may participate. It no longer governs.

BFC was my first build: a computational paradigm that gave form to the epistemology I had been following.

Which left me with the next practical question: what would I have to build within BFC for fComputation to become operative?