Cognitive offloading
Cognitive offloading
One-line summary: Using physical action and the external environment to reduce mental load is one of the best-evidenced compensations for limited working memory — but it has a catch (you forget what you offload), which is exactly why a capture system must design retrieval in.
The insight
Cognitive offloading (and its prospective-memory special case, intention offloading) is the cognitive-science construct underneath externalize-and-resurface. The evidence is strong and convergent: external memory aids are the single most effective prospective-memory intervention, offloading reduces inter-individual performance variability, and the benefit grows with memory load. The decisive twist for system design is the "Google effect": offloading an item tends to make you forget that very item (it behaves like a "forget" instruction), while improving memory for other, non-offloaded items ("saving-enhanced memory"). So once you externalize, you genuinely depend on the external store — which makes reliable resurfacing non-optional, not a backup.
Evidence
- From 2026-06-07-academic-research-adhd-pi-strengths-and-externalized-cognition: in a meta-analysis of 73 prospective-memory intervention studies, external memory aids were the most effective category (Hedges' g ≈ .805) and the most ecologically valid; combination interventions did worse ("more is not always better") (Jones et al. 2021).
- From 2026-06-07-academic-research-adhd-pi-strengths-and-externalized-cognition: intention offloading is "highly effective," metacognition-guided, with stable individual biases (Gilbert et al. 2022); it reduces inter-individual variability and helps prospective more than retrospective memory (Burnett et al. 2025).
- From 2026-06-07-academic-research-adhd-pi-strengths-and-externalized-cognition: the "Google effect" / saving-induced forgetting — offloaded items are themselves forgotten while other items are remembered better (Gilbert 2024; Kelly et al. 2019).
- From 2026-06-07-academic-research-adhd-pi-strengths-and-externalized-cognition: reminders help most under high load, and target-action reminders work while target-only or action-only do not — the reminder must carry both the cue and the action, at no cost to the ongoing task (Peper & Ball 2022).
Design implications
- For sparks: capture the next move alongside the idea (target-action), and treat resurfacing as a hard requirement — the Google effect means an un-resurfaced spark is lost from both you and the system.
- The system earns the most under high cognitive load — i.e., exactly when an ADHD mind most needs it.
Contradictions / tensions
- Most offloading evidence is from general/aging/clinical-non-ADHD samples; ADHD-specific reminder RCTs are sparse (see adhd-resurfacing-cadence).
- Clinical populations can offload non-optimally due to metacognitive miscalibration (e.g., under-scaling reminder use as difficulty rises) — whether ADHD over- or under-offloads is open.