Location and binding in visual working memory
DOI: 10.3758/bf03195932
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study investigates the organization of visual working memory (VWM), specifically addressing whether spatial locations are stored separately from object features or integrated into bound object representations. Motivated by conflicting evidence regarding the separation of spatial and object stores, the authors test the hypothesis that locations play a central role in the encoding and maintenance of bound objects, even when spatial information is task-irrelevant. The research aims to determine if feature binding is automatic upon attention and how dependent these bindings are on spatial coordinates over time. The researchers conducted five experiments using change detection paradigms. Participants viewed displays of three colored shapes or letters, followed by a retention interval and a probe display. In Experiments 1 and 2, participants judged whether features matched the original display, ignoring changes in binding or location. In Experiments 3, 4, and 5, participants judged whether the bindings (feature pairings) remained the same. The design manipulated whether probes appeared in original or new locations and whether they were presented as whole displays or single items. Experiment 3 additionally varied retention intervals (100 msec vs. 900 msec) to assess the consolidation of bindings. Articulatory suppression was employed in some conditions to prevent verbal coding. The results demonstrated that features are spontaneously integrated with locations when attended, even if binding is irrelevant to the task. In feature recognition tasks, changing the binding impaired performance, particularly when probes appeared in original locations, indicating that feature-feature bindings are mediated by feature-location links. However, overall feature recognition was not significantly impaired by location changes alone, suggesting features can be accessed from separate stores. Conversely, memory for bindings was highly vulnerable to location changes. Performance on binding tasks dropped significantly when probes appeared in new locations, especially at short retention intervals. This effect diminished slightly with longer delays, suggesting that direct feature-feature links consolidate over time, though they remain dependent on spatial context. Single-probe tests reduced the impact of location changes compared to whole-display tests, implying that relative spatial patterns support binding retrieval. The findings challenge the view that spatial and object information are held in entirely distinct, independent stores in VWM. Instead, the study concludes that locations are integral to the early encoding and maintenance of bound objects. While features can be retrieved independently of location, the binding of features into coherent objects relies heavily on spatial coordinates. This suggests that VWM representations are not purely modular; rather, attention leads to automatic integration of features and locations, with spatial information serving as a critical anchor for object identity and continuity.
Key finding
Visual working memory holds features in distributed feature maps that survive location changes, but binding of features into integrated objects is location-dependent: changing locations selectively impairs binding memory while leaving feature memory largely intact.
Methodology
lab_experiment
Sample size: 16
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28 (4 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 2 | 2026-06-02 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | failed | — | — | — | 4 | 2026-07-02 |
| promote | success | — | — | — | 2 | 2026-06-06 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 23 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.