Response to Commentaries

My thanks, first, to the commentators for taking the time and effort to respond.

Empiricism and quantification

The question of empiricism and quantification ran through the commentaries as a common, and generally positive, theme. Almost all the commentators appeared to welcome the fact that evidence generated in a psychoanalytic context should be tackled in a manner that suggested a desire to measure phenomena in a tractable manner, and with a reasonable level of experimental control. There was, of course, some disagreement about the specifics of how this was achieved (e.g., Blechner, DeLuca, Schnider, and others). For example, several commentators raised, quite rightly, matters related to sampling and control in the study: why this series of patients, and which criteria were employed in selecting them; why were the patients not worked up in greater detail in terms of their performance in conventional neuropsychological assessment, and their everyday behavior not documented in greater detail; why were there not more appropriate control groups?

Answers to these questions follow, in large part, from an understanding of the nature and purpose of the study. The study was conducted post hoc, as a reinvestigation of previously studied cases. Our goal was not to investigate confabulations in the manner expected of an entirely conventional neuropsychological study. This would have required a more systematic neuropsychological work-up of the patient, a betterdesigned method for ensuring that all possible confabulations were evaluated, and the addition of a control group. Such a study would also be submitted for publication in a conventional neuropsychological publication—as we have now done (e.g., Fotopoulou, Solms, & Turnbull, 2004; Turnbull, Berry, & Evans, 2004). These studies have addressed basically the same question as that in the study under consideration here, and they produced a more or less identical conclusion of possib