Language is one of the hallmarks of intelligence, demanding explanation in a theory of human cognition. However, language presents unique practical challenges for quantitative empirical research, making many linguistic theories difficult to test at naturalistic scales. Artificial neural network language models (LMs) provide a new tool for studying…

Computational models have shown that purely statistical knowledge about words' linguistic contexts is sufficient to learn many properties of words, including syntactic and semantic category. For example, models can infer that "postman" and "mailman" are semantically similar because they have quantitatively similar patterns of…

Syntactic bootstrapping facilitates children's initial learning of verb meanings based on syntactic information. A challenging case is the argument-drop languages, where the number of argument NPs is not a reliable cue for distinguishing between transitive and intransitive verbs. Despite this fact, the availability of syntactic bootstrapping in…

Definite expressions may be used to introduce a referent into discourse when their familiarity between speaker and listener can be inferred, a strategy known as bridging. However, for a number of reasons, bridging may be difficult to acquire compared to the acquisition of indefinite introductions for noninferable referent types, with the native…

Eight experiments evaluated a core assumption of several theories of text processing, the shared resource assumption, which states that component text processes share limited processing resources. Short texts each contained two critical sentences that together warranted a causal inference. The syntactic structure of the second sentence was either…

Analyzed "if" sentences in conversations in the homes and preschools of four-year-old children. Parents and teachers used "if" more often than did children. Children and parents did not differ in the proportion of "ifs" that had the linguistic properties of a conditional premise. (BC)