Catching the brain in the act of confabulation: a fMRI study

We explain our decisions with ease, but little are we aware of how wrong these explanations can be. A growing body of research has shown that we sometimes unknowingly invent plausible but inaccurate stories (i.e. we confabulate) when accounting for our behaviour in everyday life. This led several theories to postulate that confabulation and introspection rely on very similar interpretative neurocognitive mechanisms generally used for social cognition, especially in the left hemisphere. To test this hypothesis, we conducted the first neuroimaging study of confabulation in the normal population with fMRI. Using the choice blindness paradigm, we led participants to believe that they chose faces that they actually did not choose as being the most attractive and to justify their purported decisions (i.e. to confabulate). Our results showed that confabulation was associated with the right side of the mentalizing network typically involved in social cognition and self-processing (right temporoparietal junction, medial prefrontal cortex, precuneus). Confabulation also recruited areas related to reality and memory monitoring and executive functions (right rostral and dorsolateral prefrontal cortex). However, non-confabulation did not appear to recruit any other areas. We discuss how these results suggest that confabulation and introspection share the same interpretative basis, although confabulation may require a more effortful rationalization. We also discuss how the misinformation involved in our paradigm may explain executive and memory monitoring activations. Finally, we describe how our right-sided activations challenge the left-brain interpreter theory of confabulation and suggest new research tracks to study the commonalities between introspection and confabulation further.