The acquisition of Principle B

Abstract

This dissertation investigates the acquisition of pronominal interpretation in child language, with a particular focus on the Delay of Principle B Effect (DPBE), a phenomenon whereby children allow local interpretations of pronouns in contexts where adult grammar does not. While the DPBE has been widely documented in English and other languages, its underlying causes remain debated. Furthermore, its absence in many languages also remains largely unknown, with previous accounts correlating its absence to the availability of clitics.

Across three empirical studies, I examine the comprehension of pronouns and reflexives in English-, Hungarian-, and German-speaking children, combining an offline judgment task with an online eye-tracking method. The findings reveal clear cross-linguistic variation: English speaking children display a robust DPBE, whereas Hungarian- and German-speaking children do not. Importantly, however, eye-tracking data from German show that pronouns present greater difficulties than reflexives in both children and adults, suggesting that pronouns are harder to integrate, even when the DPBE is not observed.

Taken together, these results challenge previous accounts that attribute the DPBE to performance factors or maturational constraints on the use of domain-specific rules. The results of the German and Hungarian study further challenge previous explanations of the cross linguistic asymmetry, as the absence of clitics in the two languages would predict the DPBE. Instead, I argue that children’s non-adultlike interpretation of pronouns is due to an initial mis-representation of the pronominal paradigm. This can occur in languages such as English, where the morphology of pronouns and reflexives is less distinct compared to, for example, German and Hungarian. I further argue that reanalysis of the paradigm depends on the ability to compute pragmatic inferences. Specifically, children must learn that the choice between a pronoun and a reflexive encodes a contrast in meaning.

This account predicts that in languages lacking a clear morphological distinction between reflexives and pronouns, children are expected to show prolonged non-adultlike interpretations of pronouns. Moreover, within languages where the DPBE is observed, improvements in pronoun comprehension should correlate with the development of broader pragmatic skills. Specifically, children’s ability to correctly interpret pronouns in sentences such as John saw him (with him̸= John) should correlate with their ability to compute scalar implicatures, such as interpreting John ate some apples as “some but not all”. On this view, both phenomena depend on the development of Gricean pragmatic reasoning about alternatives: during the acquisition of the pronominal paradigm, children are expected to come to reason that if a speaker uses himself rather than him, it is because reflexives carry a more specific meaning than object pronouns. This ability is in turn likely to relate to measures of Executive Functions and Theory of Mind, as suggested by the literature on the acquisition of scalar implicatures. This opens the door for future research comparing acquisition of different phenomena (scalar implicatures, pronouns)that may be affected by a common underlying set of cognitive functions, whose development has implications for language acquisition.

The dissertation thus contributes to the debate concerning the locus of children’s ungrammatical pronoun interpretation. Empirically, it expands the cross-linguistic picture by providing a close comparison of children’s behaviour across three languages using the same task. Theoretically, it advances an account in which the DPBE relates not to strictly syntactic constraints, but rather to morphological mis-analysis during the early years of language acquisition.

Date of Award	Mar 2026
Original language	English
Supervisor	Raffaella Folli (Supervisor), Christina Sevdali (Supervisor) & Juliana Gerard (Supervisor)