Cross-linguistic structural priming effects suggest that bilinguals have shared or connected memory representations for similar syntactic structures. This predicts an influence of the production preferences of one language in the other language (Bernolet & Hartsuiker, 2018). We hypothesized that shared structures will lead to a facilitatory effect on production frequencies, whereas connected structures may sometimes lead to an inhibitory effect due to competition between structures. We compared the production preferences and priming effects in Dutch for the frequent by-phrase-final and the uncommon by-phrase-medial passive between Arabic/Berber–Dutch and Turkish–Dutch heritage speakers and native speakers of Dutch. Arabic/Berber–Dutch speakers produced more agentless passives –that is, the alternative shared between their two languages. In contrast, Turkish–Dutch speakers produced less by-phrase-medial passives, although these are less uncommon in Turkish. This inhibition effect suggests that syntactic structures may sometimes be connected rather than shared, although the exact mechanisms behind the inhibitory effects require further research.

跨语言结构启动效应表明，双语者对相似的句法结构具有共享或连接的记忆表征。这预测了一种语言的制作偏好对另一种语言的影响（Bernolet & Hartsuiker，2018）。我们假设共享结构会对生产频率产生促进作用，而连接结构有时可能由于结构之间的竞争而导致抑制作用。我们比较了阿拉伯语/柏柏尔-荷兰语和土耳其-荷兰语传统使用者和荷兰语母语使用者之间频繁出现的词尾被动和不常见的词中被动的产生偏好和启动效应。说阿拉伯语/柏柏尔语-荷兰语的人会产生更多的无代理被动语态——即两种语言之间共享的替代语。。这种抑制效应表明句法结构有时可能是相连的而不是共享的，尽管抑制效应背后的确切机制需要进一步研究。