Psychologists who carry out research with human subjects, especially young children, have special responsibilities to consider the ethical implications of their work. In this regard, Herrmann and Yoder raise a number of important issues about the potential consequences of continued research on a topic of considerable signi®cance for both basic and applied analyses of children's cognitive development. More speci®cally, they argue that research on the modi®ability of memory is problematic because young children may experience negative consequences from participating in studies that result in the establishment of false memories. Herrmann and Yoder maintain that the risks to children who take part in these investigations are so great that this line of inquiry should be discontinued forthwith. Our analysis of the literature, however, leads us to quite contrasting conclusions. Indeed, approaching the broad question of memory modi®ability from both clinical (BNG) and developmental (PAO) perspectives, we take issue with Herrmann and Yoder's assessment of the extent to which children are at risk and hence of the dangers of sustained research on this topic. We ®nd their alarmist tone to be both disturbing and counterproductive and suggest that discussions of potential risk to individual research participants be coupled with assessments of the bene®ts that may accrue from continued programmatic research.
Herrmann and Yoder's stimulating article thus prompts us to re¯ect on the costs and bene®ts of studies dealing with `implanted' memories in children. And in carrying out this cost±bene®t analysis, we ®nd it important to pose questions with reference to at least three levels of analysis: ®rst, the individual research participants; second, the discipline of psychology, and ®nally, society at large. Thus, without favouring any procedure that puts a child at risk, it is essential to examine carefully the evidence for claims of risk to the children who take part in studies of the malleability of memory. Moreover, it is equally important to keep in mind that information obtained from these studies may contribute substantially to our basic understanding of memory processes and, further, may be of critical importance for the well-being of other children and their families who ®nd themselves involved in real-world legal proceedings.
This type of analysis of the risks and bene®ts of research on memory implantation also leads us to ask questions about what it would be useful to know about memory modi®ability and about what we actually know at this point in time. Ideally,