What Computation Cannot Be: Linguistic Colonization, Formal Limits, and a Taxonomy of Irreducibly Human Faculties

Abstract

The vocabulary of artificial intelligence has colonized human self-description. Terms historically reserved for conscious agents—intelligence, learning, understanding, creativity—now routinely designate computational processes that bear no phenomenological resemblance to the experiences that they originally named. This conceptual erosion is not merely semantic; it undermines our capacity to articulate what distinguishes human cognition from machine operation at a moment when that distinction matters most. This article develops a dual-path argument for the identification of irreducibly human capacities. A via negativa assembles four independent lines of formal and philosophical reasoning—Gödel’s incompleteness theorems, Turing’s halting problem, Searle’s Chinese Room, and the frame problem—that establish principled limits to what computation can achieve. A via positiva draws on phenomenology to positively characterize the structures of human existence that lie beyond those limits: embodiment, intersubjectivity, temporality, being-in-the-world, and subjective experience. The convergence of these independent paths—each pointing from its own starting point toward the same boundary—grounds a taxonomy of five Unique Human Faculties (creativity, moral judgment, self-determination, empathy, and sentience) that resist mechanization in principle. The article concludes that the burden of proof now lies with those who claim that formal computation can exhaust human understanding, and argues that the simulation–possession distinction provides a principled basis for restoring conceptual clarity to discourse about artificial intelligence.