Item

Abstract

While sentence processing remains generally well preserved with increasing adult age, difficulties arise when sentence processing taxes verbal working memory. Such difficulties may be related to age differences in the use of syntactic and/or semantic associations to reduce the memory load (Stine-Morrow & Payne 2016). Both syntactic and semantic associations enable the grouping of single words into larger units (Bonhage et al. 2014). Therefore, in this behavioral study, we varied the availability of syntactic and semantic associations. Syntactic associations were made available in sentences, which were contrasted to word lists; semantic associations were made available in meaningful sentences and lists, which were contrasted to pseudoword sentences and lists (Table 1). We evaluated the extent to which older compared to younger adults may differentially use the availability of syntactic or semantic associations to cope with verbal working memory limitations. Eased by the availability of syntactic or semantic associations, participants judged whether the serial order of two words from either sentence or list matched the order in which they were previously encountered. Varying the level of verbal working memory demands, two experiments were conducted. In experiment 1, 8-word stimuli were used; in experiment 2, 11-word stimuli were used, increasing verbal working memory demands. 27 younger (mean age: 26 years) and 26 older adults (mean age: 66 years) participated in experiment 1. Similarly, 26 younger (mean age: 25 years) and 27 older adults (mean age: 64 years) participated in experiment 2. While the effective use of syntactic associations was expected to decrease with age, the use of semantic associations was hypothesized to be enhanced for older compared to younger adults (Stine-Morrow & Payne 2016). The results showed that, only when verbal working memory demands were high (i.e., in experiment 2 when using longer stimuli), the use of syntactic regularities was indeed compromised at old age (experiment 2, sentence structure x age group interaction, F(1,51) = 9.88, p < .01), while the benefit of semantic information for sentence processing was comparable across age groups (experiment 2, semantic information x age group interaction, p > .05). In light of the reduced use of syntactic associations, our findings may suggest that semantic information processing may become relatively more important for successful sentence processing with advancing adult age, possibly inducing a syntactic-to-semantic-processing strategy shift.