Metacognition

Metacognition is cognition about cognition — the capacity of a cognitive system to represent, monitor, and regulate its own cognitive processes. When you notice that you do not understand something, decide to re-read a paragraph, recognize that your memory of an event is unclear, or judge that a plan is likely to fail before executing it, you are engaging in metacognition.

The concept originates with developmental psychologist John Flavell, who distinguished metacognitive knowledge (what one knows about cognition in general and one's own cognitive processes in particular) from metacognitive monitoring (ongoing awareness of one's current cognitive states) and regulation (using that awareness to adjust behavior). This tripartite structure has proven useful and has generated substantial empirical research, particularly in educational psychology.

The philosophical problem is that metacognition seems to require that a cognitive system be transparent to itself — that the monitor have reliable access to the states it monitors. Evidence suggests this access is systematically limited. Introspective reports of cognitive processes frequently diverge from what those processes actually compute, and the divergence is not random noise but systematic bias: subjects explain their decisions using post-hoc narratives constructed after the decision is made, rather than reporting the actual causes. This is not a pathology. It is the normal operation of the metacognitive system.

The metacognitive system does not have privileged access to the cognitive system it monitors. It has access to outputs and to certain process markers — feelings of familiarity, of difficulty, of confidence — that correlate imperfectly with underlying states. The feeling that one understands something is one such marker. Its correlation with actually understanding the thing is weaker than academic culture assumes.