Misattribution of Memory: A Real Memory Filed Under the Wrong Source

A memory does not arrive with a label saying where it came from. The source has to be judged, and the judgment is made from the qualities of the memory itself: how vivid it is, how much perceptual detail it carries, whether it fits what else you know, how easily it comes to mind. Marcia Johnson, Shahin Hashtroudi and Stephen Lindsay set this out in their 1993 source monitoring framework. Perceived events usually carry more sensory detail than imagined ones, so detail is a fair cue to having seen something. It is also a cue that can mislead, because an imagined or inferred event that was processed richly can carry as much detail as a real one.

The word list above produces the error in a controlled way. Every word on the list is an associate of chair, so chair is activated again and again while the list is shown, even though it never appears. By the time you write your words down, chair is as familiar as the words you saw, and the feeling of familiarity gets attributed to the list rather than to your own thinking. In Roediger and McDermott's experiments people did not merely guess the missing word: they said they remembered hearing it, often with the same confidence as words they had heard.

Schacter's 1999 paper on the seven sins of memory places this under misattribution, alongside two other members of the family. Source confusion is remembering a true piece of information with the wrong origin, such as a rumor recalled as a news report. Cryptomnesia, or unintentional plagiarism, is producing someone else's idea as your own because the memory of encountering it has lost its tag. All three have the same structure: the content is retained, the attribution fails.

Familiarity without a source has its own classic demonstration. Jacoby, Kelley, Brown and Jasechko had people read a list of made-up names in 1989; a day later those names were more likely to be judged famous than new made-up names, because they felt familiar and the reason for the familiarity had faded. Misattribution, on this account, is a failure of the inference that normally tells you what a memory is.

Henry Roediger and Kathleen McDermott published "Creating false memories: Remembering words not presented in lists" in the Journal of Experimental Psychology: Learning, Memory, and Cognition in 1995. They revived a procedure James Deese had described in 1959: lists made entirely of associates of a single word that is itself left out. In their first experiment, students heard six lists of twelve words each, with the missing word, such as sleep for the list bed, rest, awake, tired, dream, and wrote down what they remembered immediately after each list. The missing associate was recalled after 40 percent of the lists.

The second experiment expanded the materials to 24 lists of 15 words, built like the list on this page, and the false recall rate rose to 55 percent. A recognition test followed, mixing studied words, the missing associates and unrelated new words. Participants called the missing associates old at a rate comparable to the hit rate for words they had actually heard, and when asked whether they remembered the moment of hearing a word or merely knew it had been on the list, they often claimed to remember hearing the word that was never spoken.

The error was systematic rather than random: the one word that appeared on nobody's list was the word everybody's list pointed at, and it appeared on an immediate test, a condition under which memory for short lists is normally good. The procedure is now known as the DRM paradigm, after Deese, Roediger and McDermott.

Schacter's 1999 American Psychologist article, "The seven sins of memory," organized this and related findings into a taxonomy of three sins of omission, transience, absent-mindedness and blocking, and four sins of commission, misattribution, suggestibility, bias and persistence. He placed the word-list illusion under misattribution and argued that the same source monitoring failures explain false recognition in the laboratory, mistaken eyewitness identification and unintentional plagiarism.

The rows below are illustrations of where a retained memory can be given the wrong origin, not findings from the studies cited on this page.

The error happens at the point where a memory is judged, so the counters are habits of judgment and the records that make judgment unnecessary.

• Ask for the source before you trust the content. "Where did I get this" is a different question from "is this familiar," and the second one is the one misattribution answers for you.
• Distrust fluency that arrives without a context. An idea or a claim that feels obviously right and has no memory of an origin attached is exactly what a misattributed memory feels like from the inside.
• Keep external records where attribution matters: dated notes, saved messages, a reading list. Jacoby's participants could not tell yesterday's made-up names from famous ones; a written list would have settled it in seconds.
• Separate what you saw from what you concluded, at the time. Witnesses and note-takers who record inferences as inferences give their later selves a chance to tell them apart.
• Expect the error in confident, detailed memories as much as vague ones. In the 1995 experiments, people claimed to remember the moment of hearing a word that was never spoken.
• When a dispute is about who said what, treat both memories as reconstructions and look for a record.

Misattribution is a family name. Several pages on this site cover members of the family or its neighbors, and the differences are worth keeping.

• It is not the whole of false memory, which is the broader result that people remember events that did not happen or happened differently. Misattribution is one route to a false memory, the one in which real content is given the wrong origin; that page uses a different word list and asks a different question of it.
• It is not only source confusion, which is the member of the family in which the content is true and the origin is wrong, such as a rumor recalled as news. Misattribution also covers false recognition, where familiarity is assigned to an event that never occurred at all.
• It is not only cryptomnesia, the member in which someone else's idea is produced as your own. That page treats unintentional plagiarism on its own terms.
• It is not suggestibility, which Schacter lists as a separate sin: a memory altered by leading questions or information supplied by someone else. The word-list illusion on this page arises with no misleading input from outside.