Confabulation is usually assumed to result from a deficit in either the memory verification processes alone or in both the search and the verification processes. The present study concerns a patient who, in contrast to other patients, displayed confabulations but had preserved memory verification abilities. She exhibited only a selective impairment of the search processes. Recognition abilities were preserved, and cued recall was better than free recall. On the latter task, she recalled fewer correct items and produced more intrusions than control subjects. The patient had normal performance in several tests usually assumed to tap "executive functions." It is thus concluded that an impairment in verification, regardless of whether it is specific or not to memory, is not a necessary component of confabulations. The case is discussed in relation to two memory control processes models

The results are in favour of the hypothesis that verbal comprehension difficulties lead to the production of confabulation. They are inconsistent with the idea that memory monitoring impairment is necessarily involved.