Dual-task interventions during the reactivation of aversive or traumatic memories, including bilateral alternating stimulation used in EMDR and certain concurrent visuospatial tasks, occupy an important place in contemporary debates on the modification of emotional memories. Their immediate effects on vividness, emotionality, and recall-related distress are relatively well documented, but their mechanistic interpretation remains debated. The literature still tends to bring together, and at times conflate, distinct levels of explanation: concurrent taxation of limited working memory resources, extinction, inhibitory learning, the orienting response, prediction error, and memory reconsolidation.

The effects of dual tasks can only be understood rigorously if several levels of inference are distinguished. A reduction in vividness or distress constitutes an important immediate effect, but it does not, by itself, allow one to conclude that the traumatic memory has been durably transformed. Similarly, a decrease in emotional responding may reflect safety learning or contextual inhibition without necessarily implying modification of the original memory. Reconsolidation constitutes a strong mechanistic hypothesis, but it requires strict conditions of reactivation, novelty, prediction error, and restabilization, which are rarely established with certainty in ordinary clinical contexts.

We therefore propose a critical clarification of the levels of inference involved in interpreting dual-task interventions. This clarification makes it possible to distinguish the immediate effect on the experience of recall, increased tolerance of contact with the memory, the clinical de-actualization of traumatic memory, and the more demanding hypothesis of memory updating. De-actualization does not refer here to a new mechanism, but to the central clinical issue long described in the trauma tradition: the shift from a memory experienced as a threatening presence in the present to a memory recognized as an event located in the past and capable of being integrated into autobiographical memory.

This position leads to a reformulation of the clinical and methodological implications of the field. Clinically, it invites us not to reduce therapeutic change to the immediate decrease in subjective units of distress or image vividness, but to examine the durability of change, its generalization beyond the session, and the transformation of the autobiographical status of the memory. From a research perspective, it calls for protocols capable of separately measuring immediate working memory effects, safety learning, autobiographical contextualization, and the strict conditions required to infer possible reconsolidation. The aim is therefore not to add another explanatory mechanism, but to specify what dual tasks do, and do not, allow us to conclude.