The Development of Linear Ordering Preferences in Child Language: The Influence of Accessibility and Topicality

Abstract

When communicating with their interlocutors, adults have a robust preference to order previously mentioned (‘old’) referents in the discourse before mentioning referents that have not yet been introduced in the discourse (‘new’). But in an experimental study investigating phrasal conjuncts, 3- to 5-year-olds acquiring German exhibit a ‘new-old’ ordering preference (Narasimhan & Dimroth, 2008 Narasimhan, Bhuvana and Dimroth, Christine. 2008. Word order and information status in child language. Cognition, 107(1): 317–329. [Crossref], [Web of Science ®] , [Google Scholar]), suggesting that the ‘old-new’ preference does not have early cognitive roots but changes over the course of development. Here we ask the following: When does children's ordering preference switch from the ‘new-old’ to the adult-like ‘old-new’ pattern? And can we change 4-year-olds' ‘new-old’ preference by making the ‘old’ referent more topical in the discourse? Findings show that by the age of 9 years, children have switched their preference to ‘old-new’ (Study 1). But despite discourse encouragement to use the ‘old-new’ order, younger children continue to prefer the ‘new-old’ order (Study 2). Children's non-adultlike ordering preference is robust and persists until later childhood. Factors such as exposure to the ambient language and/or cognitive maturation are likely to play a role in the development of the ‘old-new’ order in children.