Note that in all of these studies we have only used the Dutch word without the German translation as memory cues. This reactivation procedure is based on the idea of reactivating the German translation during sleep by presenting solely the memory cue (the Dutch word). However, it is conceivable that the presentation of the whole word-pair during sleep could potentially be even more beneficial for memory: Re-exposure to the Dutch word immediately followed by its German translation might reactivate and strengthen the whole memory traces to a greater extent as
compared to re-exposure to the Dutch word alone. To test this important question, in a subsequent study (Schreiner, Lehmann, & Rasch, 2015) participants learned again Dutch words and their German translations in the evening. During subsequent sleep, one third of the newly learned words were replayed without the German translation (single cue), one third was replayed immediately followed by the correct German translation (cue + feedback) and one third was not replayed during sleep (uncued). While we could replicate our former finding that cueing of single Dutch words enhances memory performance, the benefits of cueing Dutch vocabulary during sleep were completely blocked when
memory cues were directly followed by correct feedback (Fig. 1E). In an additional group, exactly the same result pattern emerged when the Dutch word was immediately followed by a false German translation, without any further impairment in thefeedback condition. This result pattern strongly suggested that the blocking of the benefits of single word cueing during sleep does not depend on the content of the second word but more crucially on the timing of the second stimulus (Fig. 1F). We substantiated this conclusion by running a third group varying the timing and the content of the second auditory stimulus. Here, replayed Dutch cues were again followed by the correct German
translation but with a longer inter-stimulus interval (1500 ms instead of 200 ms, cue + late feedback), while others were immediately followed by a pure tone (cue + tone). We hypothesized, that if the timing of the feedback is substantial and further the content does not play any crucial role, a neutral second stimulus such as a tone should as well block the beneficial effects of cueing, while a longer inter-stimulus interval might resolve that. Additionally, one third of the Dutch group was again not replayed during sleep (uncued). Importantly, while playing a tone immediately after the Dutch word completely blocked any benefits, the memory enhancing effect of cueing during sleep re-emerged when feedback was played after an additional time delay (Fig. 1G). The result pattern clearly indicates that a undisturbed period of approximately 1–2 s after cue presentation during sleep is critically needed, in order to let reactivation-induced plasticity processes unfold their benefits on the stabilization of memory representation of newly learned foreign vocabulary.