The acquisition of sentence alternations : how children understand and use the English dative alternation.
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Many English verbs expressing transfer can be used in two different constructions, one with no preposition (Rick gave Kate a coffee) and one with the preposition to (Rick gave a coffee to Kate). Whenever speakers use such a verb, they have to choose between these two constructions. This choice is determined in part by some features of the two objects: all other things being equal, speakers are more likely to use whichever construction places a shorter object before a longer one (Rick gave a coffee to the tall and well-dressed woman standing next the the desk at the southern side of the room), an animate object before an inanimate one (Rick gave Kate a coffee), a plural object before a singular one (Rick gave Kate and Roy an espresso machine), and so on. This system of feature-based choices is established very well for adult language using language corpora and experiments, but there are fewer corpora and experimental studies of child language. Because of this dearth of data, it is unknown how children acquire this choice-making system: do they start making choices determined by only one of these features and add the others piecemeal, or do they learn the system wholesale and only tweak which features win out over others? The three experiments in this thesis are a first step in answering this question. They are designed to map out the effects of length, animacy, and grammatical number on these choices over the course of typical first language acquisition. Because animacy is less stable a concept than length and number, the first experiment measures children’s and adults’ conceptions of animacy more indirectly. The second experiment presents the same participants with sentences using give where one of the two objects has been replaced by noise, and measures which of a constrained set of options they gaze at and which they choose to fill the noise gap. This provides measures of their expectations and preferences for the length, animacy, and number of the objects in these gaps. The third experiment has participants reproduce give sentences with different combinations of animacy, number, and construction. Participants reproduce sentences that conform to their choice-making system more easily. The results of these three experiments show that children as young as four years already prefer the animate-before-inanimate order. The shorter-before-longer preference is not found in any age group when the difference in lengths is just one syllable. This evidence adds to a growing body of literature converging on the finding that choices in ordering phenomena are affected by the same features wherever these phenomena occur, throughout language acquisition as well as across languages. Data from the second experiment also substantiates the common assumption that touchscreen input and eye gaze are both closely linked to attention. This will allow researchers in the cognitive sciences to use touchscreens as an alternative to eyetracking more confidently.