|This dissertation investigates children’s comprehension of demonstratives, such as this and that in English. As deictic spatial expressions, the interpretation of demonstratives is context-dependent: a proximal demonstrative (e.g., this) picks out the entity near the speaker, while a distal demonstrative (e.g., that) picks out the entity apart from the speaker; crucially, the entity-speaker distance is determined by the speaker’s perspective, which varies across contexts. Studies have shown that children tend to be non-adult-like when comprehending demonstratives uttered by a speaker who has a different perspective from their own (e.g., Clark & Sengul, 1978; Zhao, 2007). To better understand children’s comprehension of demonstratives, this dissertation explores (i) the cognitive factors which might hinder children’s adult-like knowledge, and (ii) the language-specific factors which might improve children’s demonstrative comprehension. This dissertation first discusses Theory of Mind (ToM) and Executive Function (EF) and how the development of each may hinder children’s comprehension of demonstratives. Successful comprehension of demonstratives requires the listener to incorporate the speaker’s perspective, in which cognitive abilities may play a role. It has been suggested that children’s non-adult-like demonstratives may be related to their still-developing ToM (de Villiers, 2007) and EF (Nilsen & Graham, 2012). Two experiments directly tested this hypothesis with English-speaking and Chinese-speaking children, respectively. Both experiments utilized two demonstrative comprehension linguistic tasks, and two cognitive tasks measuring ToM and EF, respectively. The results from both experiments suggest that children’s successful comprehension of demonstratives may be related to their ToM development, but not EF. This dissertation then examines whether a language-specific morphological representation of demonstratives may interact with children’s comprehension in a way that prevents them from committing non-adult-like comprehension. Demonstratives in Mandarin Chinese are of particular interest because they typically occur with classifiers. Classifiers are semantically dependent on their associated referents; interestingly, classifiers are known to facilitate adults’ sentence processing (e.g., Hsu, 2006; Wu, Kaiser, & Andersen, 2009). Thus, this dissertation examined whether and to what extent the classifier may improve Chinese-speaking children’s demonstrative comprehension. Results reveal that the classifier semantics improves children’s demonstrative comprehension, particularly when the classifier semantics itself is sufficient to identify the referent. In sum, the results of the studies discussed in this dissertation suggest that both cognitive factors and language-specific factors play an important role in children’s demonstrative comprehension.