Early and late stages of working-memory maintenance contribute differentially to long-term memory formation

•Two behavioural experiments, using delayed-match-to-sample tasks using line drawings as stimuli, were administered.•During maintenance, we administered an interference task, manipulating the moment of it.•Finally, a long-term memory (LTM) task for the line drawing stimuli was administered.•Early interference during maintenance disproportionately affected LTM performance.•We provide evidence for the differential contribution of early and late maintenance.

The present paper investigated the role of early and late stages of working-memory maintenance, which have been suggested to differentially contribute to long-term memory formation. In experiment 1, we administered a delayed-match-to-sample task, requiring participants to remember line drawings of non-sense three-dimensional stimuli. In the delay phase, participants were either presented with a fixation cross (for 2 or 9 s) or with one of two different interference tasks, varying in visual overlap with the target. The interference task was presented 1.5, 4.5 or 7.5 s after target offset. Early interfering and early probing disproportionately affected performance on an unexpected subsequent recognition-memory task compared to later interference or probing. This was not modulated by the type of interference task. In Experiment 2, we examined whether the formation of a holistic internal code of the target may be a gradual process. An analogous delayed-match-to-sample task was administered, with interference after 0.5, 2.5 or 4.5 s after target offset. The early and middle interference condition similarly disproportionately affected performance compared to later interference. Hence, the present results support the view of a functional dissociation between early and late stages of working-memory maintenance and that early working-memory processes contribute particularly to long-term memory formation.