Remembrance of Lies Past: A Comparison of the Features and Consistency of Truthful and Fabricated Trauma Narratives KRISTINE A. PEACE 1 * and STEPHEN PORTER 2 1 Grant MacEwan University, Edmonton, Alberta, Canada 2 University of British Columbia—Okanagan, Kelowna, British Columbia, Canada Summary: The credibility of reports of victimization must be evaluated by police and adjudicators. The present prospective study investigated the features of truthful and fabricated narratives of trauma and their relative consistency over a 6-month period. Participants described both a genuine and fabricated traumatic experience on three occasions over the 6 months. The narrative features were coded at each phase and the stories’ consistency was examined at Times 2 and 3. We found that truthful trauma narratives contained more: details, contextual information, and emotional details, and were rated as more ‘plausible’, relative to fabricated traumas. The details of truthful narratives also were more factually consistent than fabricated narratives over the 6 month interval. This research offers basic information on the nature and consistency of traumatic memories, and potentially applicable information for discriminating true and false allegations. Copyright # 2010 John Wiley & Sons, Ltd. Fabricated allegations of victimization occur on a regular basis in legal settings, with estimates ranging from 2 to 40% of all allegations (e.g. Anthony & Watkeys, 1991; Kanin, 1994; Sheridan & Blaauw, 2004; Yuille, Tymofievich, & Marxsen, 1995). In the first national study to document the rate of false allegations of abuse investigated in Canada, Trocme and Bala (2005) examined data from the Canadian Incidence Study of Reported Child Abuse and Neglect. They found that more than a third of maltreatment cases were ‘unsubstantiated’ (i.e. no corroboration was found), and 4% of the cases had been intentionally fabricated. Credibility assessment in such contexts is a process fraught with error and there is no ‘Pinocchio’s nose’ for identifying liars (e.g. DePaulo et al., 2003; Sporer, 2008; Vrij, 2008). While certain professional groups (e.g. see Ekman, O’Sullivan, & Frank, 1999; Mann, Vrij, & Bull, 2004) and individuals, or ‘wizards’ (O’Sullivan & Ekman, 2004), appear to be better than chance at detecting lies (but see Bond & Uysal, 2007), both laypersons and legal professionals generally perform around chance, which can be disastrous in forensic settings (see Porter & ten Brinke, 2009). Yet, the Supreme Court of Canada concluded that common sense is sufficient for jurors to identify a deceptive witness (R. v. Marquard, 1993), as long as facial cues and demeanour are observable (R. v. B. (K.G.), 1993). Further, researchers have found that legal decision-makers often report beliefs concerning ‘valid cues’ to deception that are completely inaccurate (e.g. Porter & ten Brinke, 2010). There is little empirical knowledge concerning the features of false allegations of victimization or their consistency over time (e.g. Vrij, 2008). Despite an enormous body of research on behavioural cues to deception (see Porter & ten Brinke, 2010, for an overview), few reliably discriminate truth and lies (e.g. DePaulo et al., 2003). In forensic contexts, the features of deceptive narratives are particularly relevant given that verbal statements typically are provided as evidence. The utility of attending to language (relative to non-verbal and physiological cues) to catch liars has been supported by both empirical and anecdotal evidence (e.g. Sporer, 2008; Vrij, 2008). Criteria-based content analysis (CBCA) is the most widely studied and best- validated statement analysis approach, based on the idea that accounts derived from memory for an experienced event differ qualitatively and quantitatively from fictitious reports (e.g. Ko ¨ hnken, 2002). Research on CBCA has indicated that in the majority of field and experimental studies reviewed by Vrij (2008), the criteria are present to a greater extent in genuine versus fabricated statements. In particular, studies consistently have found that the quantity of details, unstructured production, contextual embedding and repro- duction of conversation are useful in differentiating truths and lies (Vrij, 2008). However, few studies on CBCA have examined its utility with real and false traumas. One notable exception was a study comparing CBCA characteristics of genuine vs. deceptive allegations of rape (Parker & Brown, 2000). Their results indicated that corroborated claims of rape contained more unstructured production, details, descriptions of interactions, reproductions of conversation and descriptions of both victim and perpetrator mental states. In another study, Lamb et al. (1997) evaluated interviews with alleged victims of child sexual abuse, and found that cases rated as being plausible/likely (i.e. substantive corroboration existed) contained more CBCA criteria than those that were unlikely to have occurred (and lacked corroboration). That being said, only some criteria were predictive of veracity (e.g. unstructured production, quantity of details, contextual embedding), whereas others were rated similarly across plausible/implausible accounts. Another approach to evaluating the credibility of memory narratives derives from the reality monitoring (RM) framework (Johnson & Raye, 1981). Proponents of this perspective argue that the origins of a ‘memory’ are reflected in its narrative characteristics; with internally generated (imagined) events differing from externally generated (perceived) events (e.g. Johnson, Foley, Suengas, & Raye, 1988). In particular, experienced events contain more perceptual, contextual, Applied Cognitive Psychology , Appl. Cognit. Psychol. 25: 414–423 (2011) Published online 21 April 2010 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/acp.1708 *Correspondence to: Kristine A. Peace, Department of Psychology, Grant MacEwan University, City Centre Campus, Rm 6-329H, 10700—104 Avenue, Edmonton, AB T5J 4S2, Canada. E-mail: PeaceK@macewan.ca Copyright # 2010 John Wiley & Sons, Ltd.