Modelling Exposure and Input in Language Acquisition

Journal Title
Journal ISSN
Volume Title
Acquiring the focus-sensitive particle 'only' requires children to integrate information from multiple different domains. Specifically, they must (i) segment and store sound forms and come to understand that constituents in Focus typically bear focal accent; (ii) figure out the linear and grammatical distribution of 'only'; (iii) figure out that 'only' introduces a restriction on a set of possibilities; and (iv) construct a set of alternatives. Previous examinations of this learning problem show that children exhibit non-target-like behaviour until later stages of development (Crain et al., 1992; Crain et al., 1994; Gualmini et al., 2003; Kim, 2011; Notley et al., 2009; Paterson et al., 2003; Paterson et al., 2006; Philip & Lynch, 2000). The difficulty children have is often attributed to either lacking requisite syntactic knowledge or the inability to construct a set of alternatives. Using a novel approach, the current project aims to shed light on the issues by investigating the role of exposure and input in the acquisition of 'only'. Chapter 2 provides a theoretical analysis from the perspective of Jackendoff’s Parallel Architecture (Jackendoff, 1997, 2002a, 2017), and argues that an approach which advocates for a direct interface between language and cognition can better account for the properties of 'only'. Chapter 3 offers a critical review of the previous studies along with an overview of the learning problem and a discussion of exposure and input. Results from the corpus study in Chapter 4 show that across development, the frequency of 'only' in caregiver speech is a significant predictor of the frequency of 'only' in child speech, and that the distribution of 'only' is similar in caregiver and child utterances. Finally, results from the modelling study in Chapter 5 indicate that learners are biased toward different distributional cues in the input, and that this bias changes over the course of development as a function of the learner’s grammar in combination with evidence corresponding to particular hypotheses about the interpretation of utterances containing 'only'.
language acquisition, cognitive modelling, corpus linguistics, theoretical linguistics, Bayesian methods, information structure, focus-sensitive particles, exposure, input, language development
Hracs, L. (2021). Modelling Exposure and Input in Language Acquisition (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from