Date of Award
Campus Access Dissertation
Doctor of Philosophy (PhD)
Developmental and Brain Sciences
In order to efficiently navigate the dynamically changing world, we must continually generate mental representations of its contents, and then update those representations in the face of new information. Oftentimes this information is received in the form of linguistic input (verbal testimony from other people) rather than change we have directly observed ourselves. Research on the development of using language to update representations shows varying points of emergence: as late as 30 months (e.g. Ganea & Harris, 2010), but was recently demonstrated in toddlers as young as 16 months using a low-demand task (Ganea, Fitch, Harris & Kaldy, 2016). This thesis expands on that recent finding by presenting a series of three experiments that explore the conditions under which verbal updating occurs at this age, using location change as a test case.
Experiment 1 examined the specificity of toddlers’ predictions about location change by modifying the Ganea et al., 2016 task to include a second location. Findings demonstrated that toddlers were unable to update in this task; the role of working memory capacity and task-related factors are discussed. Experiment 2 thus repeated the original one-location updating task, extending it to a new population: language-delayed toddlers with Autism Spectrum Disorder (ASD). Despite their language deficits, toddlers with ASD succeeded in making predictions on the basis of verbal input in the one-location task. Finally, Experiment 3 examined toddlers’ ability to revoke trust by inhibiting updating in the one-location task when an informant was unreliable. In line with prior findings on the role of language in selective trust in older children (Jaswal, Croft, Setia, & Cole, 2010), toddlers continued to update despite repeated exposure to an unreliable informant.
Together, findings provide important insights into the development of updating mental representations during the second year of life. Under simple conditions (the one-location task), typically-developing toddlers, and toddlers with ASD make predictions about unobservable change that they have heard about. These predictions are persistent and unaffected by the reliability of the informant. However, this ability is limited, as increasing task complexity (e.g. the two-location task) hinders the ability to update expectations at this age.
Fitch, Allison, "Exploring Toddlers' Use of Language to Update Mental Representations: Evidence from Children with and without Autism Spectrum Disorder" (2017). Graduate Doctoral Dissertations. 456.