Telling More Than We Can Know: Verbal Reports on Mental Processes

Richard E. Nisbett and Timothy DeCamp Wilson were both at the University of Michigan when this was published, Nisbett at the Research Center for Group Dynamics within the Institute for Social Research. Nisbett became a major figure in social cognition and the psychology of judgment, later known for cross-cultural work in The Geography of Thought; Wilson went on to write Strangers to Ourselves on the adaptive unconscious and to do influential work on affective forecasting. This paper is the one most attached to both names, and one of the most-cited articles in the history of psychology.

It appeared in 1977 in Psychological Review, the American Psychological Association's flagship theory journal — a peer-reviewed venue whose articles are arguments rather than single experiments. Nisbett and Wilson review a body of studies in which people confidently explain their own choices while the experimental design shows the real cause to be something they never mention: a position effect on which stockings they preferred, an irrelevant detail that swayed a judgment. From the pattern they draw the general claim that people have little or no introspective access to the processes behind their responses, and that their explanations are plausible reconstructions built from implicit theories about what ought to have caused the behaviour.

The piece sits as a primary review and the foundational statement of the introspection-limits and confabulation literature. It anchors the human side of this topic, and it stands with Haidt's account of moral reasoning as after-the-fact justification and Gazzaniga's left-hemisphere interpreter as the three principal demonstrations that human self-report is unreliable about its own causes. It is the empirical content behind Lipton's aside that the processes by which humans decide and the processes by which they explain may be distinct, and the human precedent for the unfaithful machine explanations Turpin would document.

The authors' stake was scientific and reputational. The paper was a deliberately provocative challenge to the standing of verbal self-report as psychological data, with immediate methodological consequences for any research that asked people why they did things. There was no commercial or ideological interest in the claim; its force was felt as a problem the discipline had to absorb rather than as a position sold to anyone.