Conscious and Non-conscious Intelligence

Abstract

Although the dominant theoretical frameworks for human intelligence—including Carroll's (1993) three-stratum model, Neisser et al.'s (1996) task force report, Gardner's (1983) multiple intelligences, and Sternberg's (1985) triarchic theory—neither assert that consciousness is necessary for intelligence nor formally recognize non-conscious processes as constitutive of it, the operationalization of intelligence has historically been confined to consciously reportable cognitive performance. This conceptual argument article argues that this confinement represents a conceptual gap rather than a principled boundary, and provides a formal argument—grounded in the American Psychological Association (n.d.) definition's functional criteria, the history of psychometric methodology, and functionalist accounts of mental properties—for why reportability constitutes a measurement proxy rather than a constitutive constraint on the intelligence construct. Both intelligence and consciousness are operationalized using American Psychological Association (n.d.) definitions, with the conscious–unconscious distinction further specified through Dehaene et al.'s (2011) Global Neuronal Workspace (GNW) model; the adoption of GNW is justified specifically within the functional-behavioral orientation shared by both definitional anchors, and does not require adjudication among competing theories of consciousness. The analysis identifies six behavioral phenomena—sleep-based memory consolidation, infant statistical learning, motor adaptation, touch typing, automatized language use, and sleepwalking—that satisfy the functional criteria for intelligence while occurring outside conscious awareness, and that span five independent research domains previously unconnected within any single intelligence framework. On this basis, a two-component framework is proposed distinguishing Conscious Intelligence (CI) and Non-conscious Intelligence (NCI) as parallel and independent systems whose interaction is characterized by five functionally distinct modes: cooperation, preemption, independent operation, enabling, and training. This framework departs from prior accounts—including Allen and Reber's (1998) descriptive characterization of unconscious intelligence and dual-process theory's default-interventionist architecture (Evans & Stanovich, 2013)—not by adding a non-conscious component to an otherwise intact model, but by rejecting the evaluative hierarchy that positions conscious cognition as the normative standard against which non-conscious processing must be measured. NCI is developmentally and evolutionarily prior to CI, functionally independent of it, and in specific domains produces outcomes that CI intervention predictably degrades rather than improves. Three framework-specific empirical predictions are derived that are uniquely generated by the CI/NCI architecture and have no natural location within dual-process theory.