Additive particles in adult and child Chinese

Abstract

This dissertation contains two parts: the theoretical analysis of Mandarin additive particles: hai, you and ye; and children’s acquisition of these particles. Mandarin additive particles are investigated with respect to the information structure and their interaction with prosody. The major claim is that the various uses of these particles are related by virtue of their interaction with information structures. When they are followed by an accented focus in their c-command domain, a default focus structure arises, and a proposition with different entities from the presupposed one is added in the discourse. When the particles associate with a preceding accented focus, two possibilities arise: pure addition of contrastive topic, or scalar addition. If the particles are accented themselves, they function as focus operated by an ASSERT operator, and they are also scope particles with different associates. Mandarin additive particles are claimed to have different scopes and associate with respect to their interaction with prosodic prominence. Stressed hai and stressed you have wide-scope which is co-extensive with the whole VP, whereas stressed ye and unstressed particles have a narrow scope of the accented constituents. These particles have different presuppositions in these conditions. The acquisition part explores the production and comprehension of these particles. A series of experimental studies have been conducted to explore 2- to 8- year-olds’ production and understanding of different variants of Mandarin additive particles. The experiments of production are mainly elicitation tasks with picture description and video description, and for children under 3, a game play is designed to create a natural environment; the comprehension tasks are act-out tasks and truth-value judgment, and children are asked to give explanations to their judgments. The results show that like other languages, Mandarin additives are acquired as early as 2 years old for all the six variants. However, the comprehension pattern split w.r.t. different variants: children comprehend stressed hai and stressed you as early as 2, though they cannot interpret stressed ye and unstressed additives correctly until 7 years old, and a few children wrongly interpret these particles as stressed hai. Our experimental results seem to suggest that children at very young age have mastered the linguistic knowledge necessary to produce correct sentences with additive particles, such as that of scope, focus, c-command relation and even the idiosyncratic properties of different particles. However, children as old as 7 could not understand ye as well as the additives with a focus following. The incapability is caused by different presuppositions of stressed ye and stressed hai and you. To understand additive particles, children have to accommodate some presupposition to the context. Stressed ye presupposes a proposition which contains a different element from the host sentences of ye, and so is that of unstressed additives with a focus following. Following Crain & Thornton (1998), I assume that presenting sentences of stressed ye without its anaphoric presuppositions will render children’s failure of accommodating correct presuppositions. The reason why children could do much better in the stressed hai and stressed you cases is that their presupposition is anaphoric within the host sentence, and children need not accommodate a presupposition with different elements. In production, the presupposition is provided in the context, and children need not accommodate the presupposition, and so they can do well in the elicitation task. It is further observed in our study that children tend to interpret stressed ye and unstressed additive particles as stressed hai. The default interpretation of sentences with additive particles is the wide-scope reading. This pattern conforms to the “principle of parsimony” proposed by Crain et al. (1994), which states that if the referential presupposition is outside the context, the parser will make the fewest assumptions about information that is not given in the sentence. Children adopt the same processing strategy as adults, and it also provides evidence to the “modularity matching model” of language processing theory.