What is the nature of the relation between a verb and its arguments? In this dissertation, I look to evidence from language acquisition for answers. Any theory of ditransitives must explain certain structural asymmetries noted for both double-object (DO) datives (e.g. Alfonso gave Derek the bat) and prepositional datives (Alfonso gave the bat to Derek) (e.g. Barss & Lasnik 1986) as well as subtle but persistent meaning differences distinguishing the two dative constructions in many languages. A particular approach to argument realization, Harley (2002), does both. On Harley's approach, structural asymmetries arise from the hierarchical nature of the dative verb phrase, in which the first dative object asymmetrically c-commands the second in both constructions. In addition, the semantic facts fall out from the presence of primitives encoding possession in DO-datives (HAVE) and location in prepositional datives (LOC) that are embedded in these syntactic representations. I show that the structural asymmetries and meaning differences that have been observed for adults obtain for children as well, confirming Harley's general approach. Concerning the structural asymmetries, a series of experiments using the Truth Value Judgment task reveal that four-year-olds already have hierarchical representations within the dative verb phrase, much as adults do. This finding is based on converging evidence from Principle C and quantifier-variable binding in English and quantifier-variable binding in Kannada. The Kannada data in particular suggest that c-command (not linear order) guides children's interpretive preferences. Moreover, concerning meaning differences, a large-scale corpus study reveals that two-year-old English-speaking children demonstrate awareness of distinct possessional and spatial meaning in DO-datives and prepositional datives, respectively, in their earliest productions These results add to the considerable body of work illustrating the abstractness of children's early linguistic knowledge. I argue that the dative representations that children evidently have are not learnable if learning is construed inductively as the building up of rules and structures based solely on cues present in the input. Rather, the available evidence appears to favor deductive learning, whereby children are led to discover innately specified syntactico-semantic structures as a result of careful observation of what datives mean

Last modified

• 08/14/2018

Creator

• Joshua Baird Viau

DOI

• https://doi.org/10.21985/N27M7J

Subject

• Linguistics

Keyword

• possession
• ditransitive
• argument structure
• dative
• spatial
• acquisition

Date created

• 2007-09-05

Resource type

• Dissertation

Rights statement

• In Copyright