magazine issue 7 2016 / Issue 31

Was that how it happened? Shaping our memory for personal experiences in conversation with others

• written by Joanne Rechdan, Melanie Sauerland, Lorraine Hope & James Ost
• edited by Emir Efendic

The effects of repeated retellings on subsequent recall: Williams told stories about his experience many times

Having told his story to many different people on many occasions, Williams’ account is likely to have been gradually edited and altered.As noted earlier, speakers edit the details they discuss to serve their objectives, and the interests and assumed prior knowledge of the listener. Thus, people normally recall more about an event than they recount (Pasupathi, 2001).A major finding with respect to the selective reporting of memories in conversation is that details that go unreported, and therefore unrehearsed, may be left out of future retellings of the same memory. This phenomenon, termed retrieval induced forgetting, is thought to occur because selective retellings strengthen memories for mentioned details, while allowing memory traces of unmentioned details to weaken and decay (Anderson, Bjork, & Bjork, 1994).

Brian Williams gave several documented accounts of his ordeal over the skies in Iraq. He probably gave several more undocumented, and informal accounts—discussions with friends, family and peers—of his experience. When telling his story, Williams likely capitalized on the parts that elicited the most interest from his audience, omitting the more mundane details. Each of his retellings had the potential to alter his memory for the event, slowly forming the account he eventually related to Letterman. In fact, the gradual change in his story is reflected in a blog post he wrote in 2008. In the post, which was intended as a tribute to a war veteran, Williams wrote that the helicopter flying ahead of his was attacked from the ground, and the four helicopters in pursuit took fire. Five years later, Williams escalated from being in a helicopter that ‘took fire’ to being in the actual helicopter that was hit.

Source Misattribution: Williams could have muddled his account with that of others

Inaccuracies in co-constructed memories can sometimes be the result of source monitoring errors. Source monitoring refers to one’s ability to identify the origin of remembered information (Johnson, Hashtroudi, & Lindsay, 1993). After recalling an event with others, individuals may fail to identify inaccuracies in the account introduced by other group members, and instead attribute these to the experienced event. For example, Meade and Roediger (2002) showed a participant and a confederate pictures of typical household scenes (e.g. kitchen, bedroom, desk). Afterwards, the participant and confederate were asked to verbally recall items they had seen. In this phase, the confederate named some items that had not actually appeared in the scenes. Later, when participants completed an individual recall test for items from the scenes, they sometimes incorporated the items that had been mentioned by the confederate but were not actually present in the original scene. These results demonstrate the spread of a memory from one person to others through verbal interaction, termed social contagion of memory (Roediger, Meade, & Bergman, 2001).

Williams may have discussed his experience with those who were present in the helicopter that was hit. It is possible that through interviews and casual conversations with service personnel who were closer to the action, Williams picked up information and details about the event, and over time, incorporated them into his memory.

The role of egocentrism: Williams brought himself closer to the action

Source monitoring errors can often have an egocentric bias. People can confuse details reported by someone else as having been reported by themselves. Hyman and colleagues had participants study word lists individually and then recall them in pairs (Hyman, Roundhill, Werner, & Rabiroff, 2014). Some of the words on the lists were studied by both members of the pair, some were studied by only one member, and some had not been previously studied by either member. After engaging in collaborative recall of the words, each individual completed a source-monitoring task in which they identified the source of each recalled word. Participants made frequent source attribution errors and these errors tended to be egocentric, in that participants mistakenly attributed recalled words to themselves more often than to their partners.