Typing-based maxispan generalizes motor-based maintenance beyond speech

Maxispan procedures were developed to test whether conventional span tasks underestimate working memory (WM) performance as people may not optimally coordinate distinct maintenance mechanisms under typical procedures. In the time-based resource-sharing framework, recall can improve when a small initial subset of items is stabilized via a rehearsal-based motor program while later unrehearsed items are maintained via attention-based refreshing. Consistent with this assumption, maxispan studies have provided two key findings: improved recall for both rehearsed and unrehearsed items. Nevertheless, because prior maxispan demonstrations implemented the rehearsal component through (sub)vocal articulation, it remains unclear whether the motor-based contribution is speech-specific. We introduce a typing-based maxispan that replaces vocal rehearsal with manual motor control. Participants studied lists of digits (nine in Experiments 1 and 2; seven in Experiment 3). Critically, they cumulatively rehearsed the first four by typing them on a numeric keypad throughout list presentation and maintained the remaining digits via attention. Three laboratory experiments showed that typing-based maxispan improved immediate serial recall relative to a simple-span control, even under articulatory suppression. The advantage was driven by near-ceiling recall for the first segment of typed items and better recall for the last segment of unrehearsed items, in line with the two key findings from vocal-based maxispan studies. Our findings support a generalized motor-control account in which assigning a limited subset of items to a low-attention motor program can enhance short-term recall beyond the vocal domain. The typing-based maxispan provides a new tool for testing how motor-based and attention-based maintenance jointly determine WM span.