Telling more than we can know (Nisbett & Wilson 1977)

Claim: people have little or no direct introspective access to their higher-order cognitive PROCESSES. Asked "why", they do what an outside observer would do: apply a culturally plausible causal theory to their own case, and report its output as if it were observation. (Nisbett & Wilson 1977, Psychological Review.)

The stocking study: four identical stockings arrayed left to right; shoppers judge quality. Strong position effect — the rightmost wins about 4:1 over the leftmost. Asked why: fluent answers about texture, weave, elasticity. Asked whether position mattered: near-universal denial, some indignant. The real cause operated; the report never touched it; the reported cause was the plausible one.

Priming study: memorize word pairs like "ocean–moon"; later, name a detergent — "Tide" spikes. Asked why: "my mother uses it."

The precise thesis (sharper than "introspection is wrong"): reports are generated by theory-fitting in every case; they are CORRECT exactly when the true cause happens to coincide with the plausible theory. Accuracy is coincidence, not access.

Boundary (Ericsson & Simon): concurrent think-aloud about working-memory contents during a task is valid data; what is invalid is the retrospective "why". The boundary preserves the core.

Kin: abduction (the logical form whose human execution this exposes) · left-brain-interpreter · choice-blindness.