Reflective versus algorithmic mind
The disposition to engage reasoning, distinct from the raw capacity to do it.
What it means
Stanovich's tripartite model separates the autonomous mind (fast, intuitive defaults) from two facets of controlled cognition: the algorithmic mind, which is the computational power to sustain decoupled, effortful processing (closely tied to fluid intelligence), and the reflective mind, which is the disposition to actually initiate such override. The distinction explains why high-IQ individuals still fall for cognitive illusions: they possess the algorithmic horsepower but may lack the reflective inclination to deploy it, a gap measured by thinking-disposition scales and the cognitive reflection test. Rationality, on this view, requires both capacity and disposition, plus relevant 'mindware,' so it is not reducible to intelligence. The framework reframes many reasoning errors as failures of triggering or of available knowledge rather than of ability. It grounds the case that rationality can be taught and assessed separately from IQ.
Examples
Two people with identical intelligence diverge on a tricky logic problem because one habitually pauses to double-check intuitions and the other does not.
A quick graduate blurts that the ball costs 10p; the arithmetic is trivial for her, but nothing prompted her to stop and check the first answer that arrived.
A brilliant physicist forwards a scam email to his whole department. The capacity to spot the tell was there; the habit of pausing to ask whether it was true was not.
First described in Keith Stanovich (2009, 2011).