|Title:||The Acquisition of Japanese Case Particles and the Theory of Case Checking||Add Dissertation|
|Author:||Kazumi Matsuoka||Update Dissertation|
|Email:||click here to access email|
|Institution:||University of Connecticut, Department of Linguistics|
|Abstract:||There are several properties of Japanese grammar which make the task of language acquisition challenging for young learners,including scrambled sentences (Saito 1985) and particle omissions. Nevertheless, young children's usage of Case-particles contain few errors. (Clancy 1986) This strongly suggests that children take advantage of innate grammatical knowledge, which guides them during the acquisition of the Case system in Japanese (Chomsky 1981).
This study was conducted to investigate whether children, regardless of adult input, show a similar pattern in their acquisition of Case particles. The children's data are from three sets of corpora, independently transcribed in the CHILDES format (MacWhinney and Snow 1990). The age range of the children studied was 1;11-3;3. The frequency of the three Case particles _ga_ (the Nominative Case particle), _o_ (the Accusative Case particle) and _ni_ (the Dative Case-particle) were analyzed and compared to the adult speech in the same corpora.
It was found that the Case particle development follows a universal acquisition sequence across the three children: ga-ni/o. (The particle _ni_ was used productively significantly earlier than _o_ only in the AKI Corpus.) Data from the adult speech show that this sequence is not a result of parental input. Also, it was found that the Tense morphemes appear significantly before the first use of the Case-particles (p<<001). The data is consistent with the assumption that the Case feature on an NP is licensed by functional heads (Chomsky 1995, Koizumi 1995, Ura 1996).
To address the issue of Case-checking of the 'nominative object' in the stative predicate, the children's use of the Case-particle _ga_ with stative predicates was studied. When two arguments appear with a stative predicate, the Nominative particle _ga_ can be used for both the subject and the object in adult speech. The children used only one Nominative Case-particle _ga_. This is consistent with the argument, in Ura (1996), that the availability of multiple-checking is a parametric option.
The majority of the children's errors in the usage of the Case particles overall is the overuse of _ga_, where _ni_, _o_, or no particle is required. This observation suggests that young children assume that the Nominative particle _ga_ is used as the default Case in Japanese.
Children's early use of the particle _ni_ was also studied. It was observed that four types of _ni_ emerged early and were frequently used throughout the corpora. Those four types consist of two types of dative _ni_ and two types of postposition _ni_. These types of _ni_ are argued to be proto-postpositions, which mark the physical location or the final physical location of the item as a result of the action described by the predicate. The fact that the particle _ni_ in double object construction is acquired later than other types of the Case-particle _ni_ is discussed as possible empirical support for the existence of a grammatical property which is relevant to the acquisition of small-clause constructions (Snyder and Stromswold 1997).