Task switch costs scale with dissimilarity between task rules.
DOI: 10.1037/xge0001598
archive: archived pipeline: cataloged
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper investigates how task switch costs are determined by the similarity between task rules, challenging the conventional binary view of task switching (repeat vs. switch). The authors hypothesize that task representations are organized like cognitive maps, where the conceptual distance between two task rules dictates the magnitude of reconfiguration required during a switch. Consequently, greater dissimilarity between task rules should result in larger switch costs. To test this, the researchers developed an experimental paradigm using perceptual decision-making tasks where task rules were parameterized by the orientation of a reference arrow. By manipulating the angular difference between consecutive reference arrows (ranging from 0° to 140° in Experiments 1 and 2), they could parametrically vary task dissimilarity. The study comprised four experiments. Experiment 1 (N=53) and Experiment 2 (N=51) tested the core hypothesis. In Experiment 1, participants judged which of two test arrows was closer to a visible reference arrow. In Experiment 2, the reference arrow was replaced by a learned fractal cue to rule out perceptual confounds. Both experiments found that response time (RT) significantly increased as the angular difference between task rules increased, supporting the parametric switch cost hypothesis. Additionally, response switch costs persisted even when the same task rule reappeared after intervening trials, suggesting that each specific rule is maintained in a separate task representation rather than a single compositional one. Experiments 3 and 4 were designed to rule out alternative explanations. Experiment 3 (N=71) used a continuous range of reference arrow orientations (0–359°), forcing a single "meta-task" representation where the rule was abstract ("find the closer arrow") rather than specific. This design allowed the authors to test whether the parametric effect was driven by mental rotation or attentional shifts rather than task reconfiguration. The results indicated that the parametric switch cost in Experiment 3 was significantly smaller than in Experiments 1 and 2, suggesting that the effect in the initial experiments was not merely due to spatial or perceptual processing but reflected the reconfiguration of distinct task representations. Experiment 4 further controlled for attentional and saccadic confounds. The findings demonstrate that task switch costs scale continuously with the dissimilarity of task rules, extending the traditional binary task-switch effect to a parametric framework. This supports the theory that task knowledge is organized in a cognitive space where similar tasks are represented closer together. The results have significant implications for understanding cognitive flexibility, suggesting that the brain balances interference and reconfiguration costs by organizing task representations in a common reference frame, similar to how perceptual stimuli are organized in psychological space.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 41 | 2026-08-22 |
| extract | success | cached | — | — | 3 | 2026-08-23 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.8-27b-gittensor | summ-v5 | 2 | 2026-08-23 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.