What Psychology Cannot See, It Destroys Invisible Value Destruction in Intelligence Measurement, Far Transfer, and Meritocratic Selection

Intelligence tests measure what can be assessed quickly under time constraints—processing speed, pattern recognition, retrieval efficiency—and institutions treat these measurements as equivalent to intelligence itself. This paper formalizes the consequences of this conflation. We introduce Invisible Value Destruction (IVD): in any coupled system, optimizing within an observable subspace (VO) structurally degrades the unobservable complement (VU). Applied to intelligence, speed-based testing renders invisible a "depth dimension"—the degree to which knowledge is deeply integrated, interconnected, and available for flexible cross-domain application—that is built over decades and cannot be assessed in timed sessions. The Model-Whole Conflation Fallacy (MWCF) occurs when predictive validity within VO is taken as evidence that VO exhausts the phenomenon. We demonstrate four consequences. First, IQ largely captures precocity in childhood (developmental rate, not destination), as evidenced by the Terman study's production of conventional success but zero transformative achievement, while rejected children won Nobel Prizes. Second, far transfer has been declared rare by a field that never consulted the neuroscience of higher-order network interactions; we argue it is ubiquitous but direction-unpredictable and source-unidentifiable, hence invisible to the standard behavioral paradigm. Third, crystallized intelligence (Gc) —which maps onto depth—peaks around age 60, yet every major institution retires people at this point and selects most aggressively in the twenties, an arrangement derived from nineteenth-century military requirements, not cognitive science. Fourth, meritocracy and technocracy lock IVD in place through a structural conflict of interest: those selected by speed-based metrics become the gatekeepers evaluating those metrics' adequacy. We address counterarguments including the AI analogy and derive eight testable predictions, including that human-AI "centaur" teams will outperform autonomous AI regardless of AI capability, because the depth dimension provides the direction-setting judgment that speed-based processing cannot generate. The WEIRD critique identified the cultural narrowness of psychology's participant base; we identify the dimensional narrowness of its measurement base—and the power structure that maintains it. Submitted as a Target Article to Behavioral and Brain Sciences (BBS-D-26-00398).