Semantic relationships among objects reduce the attention required for inter-item binding in working memory

Previous research has found that inter-item binding in working memory requires more executive attention than single items, and since intra-item binding in working memory requires more object-based attention to maintain than its constituent elements, it is possible that inter-item binding may also need more object-based attention than single items. Additionally, studies have found that semantic relationships between items in working memory help to facilitate working memory, but the mechanisms by which these semantic relationships enhance working memory are not clear. This study examines whether the semantic relationships between items in working memory can reduce the attentional resources required for inter-item binding. During working memory task, participants were presented with memory items that either had or lacked semantic associations, and were required to complete an executive attention-consuming backward counting task or an object-based attention-consuming Duncan task while performing the working memory task. Experiment 1 revealed that for non-semantic pairs, the high-load backward counting task caused significantly greater impairment to inter-item binding than to single items (t(23) = 3.47, p =.002, d = 0.71, BF₁₀ = 15.39). For semantically related pairs, Bayesian evidence strongly supported the null hypothesis of no differential executive attention cost between binding and single-item conditions (p =.787, d = 0.06, BF₁₀ = 0.22). Experiment 2 showed a parallel but weaker pattern for object-based attention: non-semantic pairs exhibited higher binding costs under the Duncan task (t(23) = 3.41, p =.002, d = 0.70, BF₁₀ = 17.27), whereas semantic pairs showed only partial attenuation of these costs (p =.123, d = 0.33, BF₁₀ = 0.59). Semantic relationships between items primarily alleviate executive attention required for inter-item binding maintenance and moderately reduced but did not fully eliminate the object-based attention demands.