The case
The owner, at U37: “I think we also may have landed on a likely (at least partial) simple, high-level explanation (or at the very least intuition) for The Doorway Effect (i.e. connect perspective and grounded experience).”
The studies, from their abstracts, all registered unread:
- The finding. Radvansky and Copeland (LIT-1392): in a virtual environment, memory for the object just picked up and carried was worse after walking through a doorway than after walking the same distance within a room, read as the updating of situation models at a boundary.
- Not the context of encoding. Radvansky, Krawietz and Tamplin (LIT-1384): the effect held with smaller displays and in a real environment, and returning to the original room did not improve memory, against an encoding-specificity account.
- Not time or disruption. Pettijohn and Radvansky (LIT-1385): travel times were similar and a 3-second delay did not change the pattern, which the authors attribute to interference at retrieval among event models that share elements.
- Not always. McFadyen et al. (LIT-1387): four experiments, in immersive virtual reality with and without a working-memory load, in video and in real movement, found no significant doorway effect; under load, false alarms to mismatched probes rose after a doorway.
- The theory behind it. Event segmentation theory (Zacks et al., LIT-1382): a rise in prediction error is perceived as an event boundary, and boundaries update the event models that guide attention and memory.
A37 also cited a 2026 virtual-reality study reporting an event-boundary effect; it was not identified and is not filed.
What it shows
A37’s reading, which the owner’s U37 prompted: “The intention hasn’t necessarily disappeared from memory. It may have lost accessibility because the current situation is being organized through a different event model, one in which the intention’s earlier contextual supports are no longer active.” So “an inaccessible intention is not necessarily an erased intention”, and information retained, information currently accessible and information deployed come apart, which is TERM-055’s distinction. That is activation interruption (CLAIM-421).
The return-to-the-room result matters most for the record’s account. A37: “So our theory shouldn’t reduce perspective to location. The relevant context is an organized, goal-sensitive relationship to the environment—not merely the current coordinates of the body.” A37 proposed two discriminating variants: a doorway crossed while a salient task continues, and a change of task with no doorway.
What it does not show
- That the effect is robust. McFadyen et al. found none under four conditions. A37: “it’s important not to propose that doorways invariably reset memory.”
- A new explanation. A37: “That isn’t yet a new scientific explanation. It’s a way of interpreting established event-cognition findings through our broader organizational-perspective framework.” The owner’s U37 offered it as at least partial, or as intuition, and the case keeps that.
- That Minsky’s K-lines (NOTE-352) or the doorway options of hierarchical reinforcement learning (THEORY-688) are its mechanism. A37 used both as analogies and said so: the first “is a theoretical account of memory, not an established neural mechanism for the doorway effect”; the second are “computational results about hierarchical action learning, not experiments on human doorway forgetting”.
- Standing. The studies are documented; the reading is the exchange’s, and the papers have not been read.
Note of 2026-10-22: what the five papers show
All five are now read. The case’s outline holds, with four corrections.
- The carried object, not the one set down. Radvansky and Copeland (LIT-1392): “the spatial shift had a larger effect on memory for the currently associated object”. The object just set down showed no reliable doorway effect. The title and first bullet are corrected. The key experiment had few trials per condition, and the rooms on either side of a doorway also differed in size and pattern (NOTE-1083).
- Interference, not loss of context. The original lab explains the effect by competition at retrieval: “The effect is a result of the competition between these two event models during retrieval, even though they are both pointing to the same response” (LIT-1385, p. 2125). Going back did not help: “memory did not improve by reinstating this context” (LIT-1384, p. 1632).
- A mixed replication record, not a failure. McFadyen et al. (LIT-1387) ran conceptual replications (passive movement in virtual reality, curtains, a different memory task), not direct ones, and their second experiment did find a doorway effect on d′, carried by false alarms: “the reduction in d’ after a shift was predominantly due to there being a higher false alarm rate for negative probes, rather than a reduced hit rate for associated and dissociated probes”. They cite independent positive studies the record does not hold. Fairly stated: well replicated within one laboratory’s set-up, mixed outside it.
- Nothing here tests an intention. The task is recognizing labelled objects. Event segmentation theory (LIT-1382) never mentions doorways, and is a theory of an observer: “the theory in its current form is a purely passive perceiver”. A37’s reading, that the effect concerns a goal-sensitive relationship rather than location, is untested by these papers; its own proposed variants (a doorway crossed while a salient task continues, a change of task with no doorway) would test it.