Bi-directional syntactic priming across cognitive domains: From arithmetic to language and back

Scheepers et al. (2011) showed that the structure of a correctly solved mathematical equation affects how people subsequently complete sentences containing high vs. low relative-clause attachment ambiguities. Here we investigated whether such effects generalise to different structures and tasks, and importantly, whether they also hold in the reverse direction (i.e., from linguistic to mathematical processing). In a questionnaire-based experiment, participants had to solve structurally left- or right-branching equations (e.g., 5 × 2 + 7 versus 5 + 2 × 7) and to provide sensicality ratings for structurally left- or right-branching adjective-noun-noun compounds (e.g., alien monster movie versus lengthy monster movie). In the first version of the experiment, the equations were used as primes and the linguistic expressions as targets (investigating structural priming from maths to language). In the second version, the order was reversed (language-to-maths priming). Both versions of the experiment showed clear structural priming effects, conceptually replicating and extending the findings from Scheepers et al. (2011). Most crucially, the observed bi-directionality of cross-domain structural priming strongly supports the notion of shared syntactic representations (or recursive procedures to generate and parse them) between arithmetic and language.