Flashbulb memories: Sorry, no Flash!

Curci and Conway (2013) concede that Flashbulb Memories (FBMs) are susceptible to all the sins (Schacter, 2001) of other memory traces. This notwithstanding they, and others, maintain that FBMs are special autobiographical memories in so far as they are associated with strong emotions at encoding, persist vividly over time and carry a high degree of confidence. Therefore, they dissent from us for having called into question the special status of FBMs (Cubelli and Della Sala, 2008) and argue their dissent criticising both our data and our account of them. We are not the first authors to have cast doubt on the specific status of FBMs (Weaver, 1993; Neisser, 2003; Talarico and Rubin, 2003). Nobody denies that there are emotionally loaded events which generate memories that appear persistent; the issue is whether the memory of these events undergoes the same processing as ordinary memories or requires separate memory mechanisms. The original metaphor accounting for FBMs as incomplete, permanent snapshots (Brown and Kulik, 1977) has been revised. In fact, considerable evidence has piled up to show that FBMs are affected by decay, interference, distortion and reconstructions no differently than memories for ordinary events (Thomson and Cowan, 1986; Crombag et al., 1996; Schmolck et al., 2000; Pezdek, 2003; Talarico and Rubin, 2007, 2009). However, the “special” party still describes FBMs as “so vivid that you remember the words that were said, smells that were present, or other apparently unimportant details such as the weather or the clothes that you were wearing” (Luminet and Curci, 2009, p. 2). Therefore FBMs are conceived as a very accurate reproduction of the learning circumstances. The “FBMs special” party maintain that since it is difficult to assess accuracy of FBMs, we should use consistency as proxy. Very consistent (emotional, long-lasting, impermeable to monitoring checks) yet false memories might present a problem to this approach. Consider for example Piaget’s visual memory of his own kidnapping which never occurred (Piaget, 1945) or Oliver Sacks’ clear recollection of his house bombing which he never witnessed (Silberman, 2003). These falsememories fulfil all the criteria defining FBMs: should they be classed as such? If FBMs refer to increased confidence of specific learning circumstances, it would be indeed interesting to identify which events generate these memories (Curci and Conway, 2013). However, this does not imply postulating special memory processing. If instead they define memories for events as frozen in time, as their description implies, this would require special memory laws and mechanisms. Hence, assessing their veridicity and plausibility is paramount. We showed that accuracy of FBMs is poor (Cubelli andDella Sala, 2008). Curci and Conway (2013) rebuff our results. The example they use from our data is that of an interviewee who was passing by in the aftermath of the explosion and later remembered only some smoke coming from the railway station coffee shop. Curci and Conway write that “it is hard to say if this detail referred to the original event or the reception context for that event” (p.??), suggesting that this participant could have merged memories from different sources. We maintained that this memory could hardly be real as the coffee shop was completely destroyed by the bombing together with the rest of the station. Whether it be a reconstruction or a false memory, either way it is inaccurate. The “FBMs special” party maintain that FBMs are a “special category” (Curci and Conway, 2013, p. ??) of memories, and that they are “more robust” than memories for ordinary events (Pillemer, 2009, p. 138). Decades of FBM research do not allow us to support these statements. FBMs surely are less accurate than explicitly or implicitly we are made to believe. An important rule of the game is to consider all the evidence.