The making of false memories
For many years researchers in cognitive neurscience have known that episodic memory does not work like a tape recorder or a computer hard drive. Recollection of events is not a simple replay from a fixed store. Rather, episodic memory (and memory in general) is today seen as a dynamic - even fragile - reconstruction process. As a consequence, errors can happen, and they do. One of these kinds of recall errors are false memories. But what are the mechanisms behind false memories? Why do things go wrong? In a paper by Lampinen et al. false memories are studied experimentally. They shed light on two special features in false memories; borrowing and vividness.
The mere existence of false memories are serious news for the use of eye witness testimonies, even for victims of violent acts such as rape. If memories cannot be treated as true, but are unstable, influenced by the context in which it is recalled, how can we make use of it at all. Vivid false memories, as described by Lampinen et al., attest that even if a person is certain about his memory about an event they can be false.
The discussion of false memories also applies to the publication of memory's "shaky trace" (PDF) a couple of years ago. It's an interesting finding that memories can be altered and even deleted at the time of retrieval. Put into more practical terms; depending on how you ask your question you will get different answers from memory. What keeps the same is that the person feels that the memory is genuine.
Compelling Untruths: Content Borrowing and Vivid False Memories
by James Michael Lampinen et al.
Journal of Experimental Psychology: Learning, Memory, and Cognition - Volume 31, Issue 5 , September 2005, Pages 954-963''
Abstract False memories are sometimes accompanied by surprisingly vivid experiential detail that makes them difficult to distinguish from actual memories. Such strikingly real false memories may be produced by a process called content borrowing in which details from presented items are errantly borrowed to corroborate the occurrence of the false memory item. In 2 experiments using think-out-loud protocols at both study and test, evidence for content borrowing occurred for more than half of the false remember judgments participants reported. The present study also provides evidence consistent with recollection rejection and distinctiveness playing a role in false-memory editing.