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.