The Primitive Matters
Why Value, Not Information, Is the Brain’s Foundational Currency
The assumption that information is the brain’s foundational currency is wrong. Not wrong in a minor, correctable way. Wrong at the level of the primitive — at the level of what the theory starts from. The Hard Problem has persisted for thirty years not because the field has lacked intelligence or rigor but because it is an inevitable consequence of the frameworks’ starting point.
Pace, J. C. (2026). The Language of Stress: The Primitive Matters. FigShare. DOI: 10.6084/m9.figshare.31953402
Overview
Theories of consciousness disagree substantially about what consciousness is, where it lives in the brain, and how it might be measured. The dominant computational frameworks nonetheless share a starting point so taken for granted that it rarely appears as an assumption at all. It appears as the obvious beginning.
That starting point is information.
This essay argues that the choice of primitive is not a technical preference but a constraint on what a theory can reach. A framework built from information can describe how a system represents the world with increasing accuracy. What it cannot do without adding something further is explain why any of that content should matter to the system holding it — and once a further scoring or reward signal is attached to otherwise neutral representation, something has to read that signal, and the problem reappears one level up.
The positive case is that value is the better primitive: that a violation registers as bad directly, rather than registering as a mismatch and then being evaluated. The essay is careful about what that argument does and does not show. It separates the architectural claim, which stands on its own, from the further identification of the required state with phenomenal experience, which is contestable and is marked as the place to press. The zombie material is framed as a claim about functional viability rather than about what can be conceived, and the essay says outright that it does not reach Chalmers' argument.
A section near the end surveys what other programs take as basic, and it supports the essay's point in a way that arguing against the computational frameworks alone does not. Damasio grounds feeling in homeostasis. Panksepp grounds affect in evolved subcortical circuitry. Enactivism grounds normativity in autopoiesis. Mark Solms — closest of all to this essay's position, and explicit that affect is where explanation bottoms out — nonetheless formalizes affect as free energy minimization, where pleasure and unpleasure are decreases and increases in expected uncertainty. Four programs, four different primitives, and in every case value is derived rather than basic.
The essay makes no priority claim about that observation and says so. What it claims is narrower: that the primitive question is a real one, that it has been answered several incompatible ways, and that value has not been one of the answers — including among the researchers who argued most forcefully that feeling comes first.