To what extent are links between musical pitch and space universal, and to what extent are they shaped by language? There is contradictory evidence in support of both universality and linguistic relativity presently, leaving the question open. To address this, speakers of Dutch who talk about pitch in terms of spatial height and speakers of Turkish who use a thickness metaphor were tested in simple nonlinguistic space-pitch association tasks. Both groups showed evidence of a thickness-pitch association, but differed significantly in their heightpitch associations, suggesting the latter may be more susceptible to language. When participants had to match pitches to spatial stimuli where height and thickness were opposed (i.e., a thick line high in space vs. a thin line low in space), Dutch and Turkish differed in their relative preferences. Whereas Turkish participants predominantly opted for a thickness-pitch interpretation-even if this meant a reversal of height-pitch mappings-Dutch participants favored a height-pitch interpretation more often. These findings provide new evidence that speakers of different languages vary in their space-pitch associations, while at the same time showing such associations are not equally susceptible to linguistic influences. Some space-pitch (i.e., heightpitch) associations are more malleable than others (i.e., thickness-pitch).

Bilingual speakers have to control their languages to avoid interference, which may be achieved by enhancing the target language and/or inhibiting the nontarget language. Previous research has provided evidence that bilinguals may use inhibition (e.g., Jackson, Swainson, Cunnington, & Jackson, 2001), which is reflected in the N2 component of the event-related potential (ERP). In the current study, we investigated the dynamics of inhibitory control by measuring the N2 during language switching and repetition in picture naming. We recorded the EEG of 30 unbalanced Dutch-English bilinguals in a cued language-switching task. Participants had to name pictures in Dutch or English depending on the cue. A run of same-language trials could be short (two or three trials) or long (five or six trials). We assessed whether RTs and N2 changed over the course of same-language runs, and at a switch between languages. Results showed that speakers named pictures more quickly late as compared to early in a run of same-language trials. Moreover, they made a language switch more quickly after a long run than after a short run. In ERPs, we observed a widely distributed switch effect in the N2, which was larger after a short run than after a long run. The N2 was not modulated during a same-language run, challenging Kleinman and Gollan (2018), who maintained that inhibition accumulates over time. Our results suggests that speakers adjust the level of inhibitory control at a language switch, but not when repeatedly naming in the same language.

Bilingual speakers have to control their languages to avoid interference, which may be achieved by enhancing the target language and/or inhibiting the nontarget language. Previous research has provided evidence that bilinguals may use inhibition (e.g., Jackson, Swainson, Cunnington, & Jackson, 2001), which is reflected in the N2 component of the event-related potential (ERP). In the current study, we investigated the dynamics of inhibitory control by measuring the N2 during language switching and repetition in picture naming. We recorded the EEG of 30 unbalanced Dutch-English bilinguals in a cued language-switching task. Participants had to name pictures in Dutch or English depending on the cue. A run of same-language trials could be short (two or three trials) or long (five or six trials). We assessed whether RTs and N2 changed over the course of same-language runs, and at a switch between languages. Results showed that speakers named pictures more quickly late as compared to early in a run of same-language trials. Moreover, they made a language switch more quickly after a long run than after a short run. In ERPs, we observed a widely distributed switch effect in the N2, which was larger after a short run than after a long run. The N2 was not modulated during a same-language run, challenging Kleinman and Gollan (2018), who maintained that inhibition accumulates over time. Our results suggests that speakers adjust the level of inhibitory control at a language switch, but not when repeatedly naming in the same language.

Musical properties, such as auditory pitch, are not expressed in the same way across cultures. In some languages, pitch is expressed in terms of spatial height (high vs. low), whereas others rely on thickness vocabulary (thick = low frequency vs. thin = high frequency). We investigated how children represent pitch in the face of this variable linguistic input by examining the developmental trajectory of linguistic and non-linguistic space-pitch associations in children who acquire Dutch (a height-pitch language) or Turkish (a thickness-pitch language). Five-year-olds, 7-year-olds, 9-yearolds, and 11-year-olds were tested for their understanding of pitch terminology and their associations of spatial dimensions with auditory pitch when no language was used. Across tasks, thickness-pitch associations were more robust than height-pitch associations. This was true for Turkish children, and also Dutch children not exposed to thickness-pitch vocabulary. Height-pitch associations, on the other hand, were not reliable-not even in Dutch-speaking children until age 11-the age when they demonstrated full comprehension of height-pitch terminology. Moreover, Turkish-speaking children reversed height-pitch associations. Taken together, these findings suggest thickness-pitch associations are acquired in similar ways by children from different cultures, but the acquisition of height-pitch associations is more susceptible to linguistic input. Overall, then, despite cross-cultural stability in some components, there is variation in how children come to represent musical pitch, one of the building blocks of music.