Taking the middle stance in the debate on the nature of traumatic memories

Abstract Purpose The discussion on the degree of similarity and continuity between more neutral memories and genuine traumatic memories lies at the core of the (at times too heated) debate on the possibility of having declarative memories for traumatic personal experiences. In this paper, we aim at taking a middle ground in the debate, by examining, albeit very briefly, clinical, neurological and behavioural data from a hopefully more objective point of view. Method In discussing traumatic memories, the first necessary step is to clarify the concept of trauma, its use and the consequences of the definition in applied areas such as the legal arena. It is not meaningful to talk about traumatic memories if trauma is defined too loosely and refers to any type of negative experience. Second, we provide a very brief overview of data deriving from both sides of the debate. Results The brief review suggests that the definition of trauma has been the object of a ‘conceptual bracket creep’, extending to events and behaviours that should not be considered trauma. This has consequences on the definition of what a traumatic memory is, hindering a productive discussion on the topic. Data from clinical observations, which strongly speak in favour of the special nature of traumatic memories, unfortunately suffer from such conceptual looseness, while neurobiological studies have adopted a more strict conceptualisation of trauma, but mainly in animal models. These studies converge in indicating that neurobiological processes involved in traumatic compared with non‐traumatic memories are different, but the effect of trauma can be both of impairing and enhancing declarative memory. Behavioural studies which oppose the special nature of traumatic memories are rigorous, but such studies lack exposure to genuine traumatic experiences. Conclusion Only by taking a more dispassionate middle ground, it becomes possible to evaluate merits, flaws and the validity of results. We suggest that the nature of traumatic memories will be better understood by accepting solid data indicating that encoding and consolidation are different in case of very strong emotionally negative events (leading at times to memory impairment, but also often to memory improvement) and by integrating these data with equally solid behavioural data. Overall, traumatic memories can be special. Research should help define specific conditions for special processes to be involved.