The study of memory has long been dominated by the structural traditio
n, and especially by the experimental analysis of mechanisms of inform
ation processing. That dominance may soon be brought to an end by the
progress of neuroscience, which offers more direct ways of studying th
e mechanisms in question. At that point functional issues may move to
centre stage. Those issues include the act of remembering and its soci
al functions, the skills and presuppositions of the rememberer, the in
teraction of those skills and presuppositions with the particular mate
rial being remembered, and the determinants of accuracy and confabulat
ion in recall.