The reference problem and how children use gesture and grammatical number to solve it

This thesis examines 2- and 3-year-olds' understanding of social cues (specifically body orientation and pointing) and grammatical cues (specifically number marking) to reference in a series of three experiments using the Intermodal Preferential Looking Paradigm. Experiment 1 examines grammatical cues. It investigates whether 2- and 3-year-olds can follow grammatical number when the potential referents belong to the same category (e.g. one knife guard versus 14 knife guards). All previous relevant studies have involved potential referents that belong to different categories (e.g. one lemon juicer versus 14 honey spoons). It is possible that this manipulation makes the task of following grammatical number easier as the categorisation component is already done. If so, previous studies may overestimate children's ability to follow grammatical number in the real world. In addition, Experiment 1 asks whether children follow the combination of nominal and verbal number marking (e.g. 'There are the zoots!') more than they follow nominal number marking alone (e.g. 'Look at the zoot!'). One previous study shows a preference for multiple number marking, but only with familiar labels and referents. A preference of this kind is referred to as the effect of cue quantity. Experiment 2 examines social cues and asks whether 2- and 3-year-olds can follow body and head orientation in a referential context. This has not previously been tested experimentally and has implications for children's ability to learn from observation. Experiment 2 also asks whether children follow the combination of body orientation and pointing more than body orientation alone. In the existing literature there are hints of this preference, but previous studies are inconclusive. The experiment therefore looks for the effect of cue quantity in both grammatical and social cues. Experiments 1 and 2 were designed as pre-tests for Experiment 3 to demonstrate that children could follow the targeted cues (body orientation and number marking) in isolation. Experiment 3 builds on these experiments by exploring changes in children's use of body orientation and number marking across development. In doing so, it aims to bring evidence to bear on a possible shift from reliance on social cues to a reliance on grammatical cues. This is achieved by comparing 2- and 3-year-olds' use of these cues in two conditions: a congruent condition (in which they point to the same referent), and an incongruous condition (in which they point to different referents). Experiment 1 showed that 2- and 3-year-olds did not follow grammatical number when the potential referents belonged to the same category. Experiment 2 showed 2- and 3- year-olds did not follow head and body orientation. Experiment 3 showed, however, that when these cues were presented together (i.e. with both pointing at the same target) children did show some sensitivity to these cues. This was only the case on singular trials (i.e. when the speaker turned to look at the individual object and produced a sentence containing a singular noun). I argue that in this condition perceptual saliency, as well as body orientation and grammatical number, pointed to the individual object being the target. I interpret this either as evidence in support of the effect of cue quantity or as evidence that 2-year-olds, but not 3-year-olds, continue to use perceptual salience as a cue to reference.