The Role of Pronoun Frames in Early Comprehension of Transitive Constructions in English

Abstract

Case marking in English in available only on some pronouns and only in some cases. It is unknown whether young children acquiring English nevertheless make use of this highly restricted marking as a cue to sentence interpretation. The current study therefore examined how 2- and 3-year-old English children use case-marked pronoun frames and constructional word order cues (actives versus passives) to understand agent-patient relations in transitive sentences containing novel verbs. In a pointing comprehension test, 2-year-olds used pronoun frames containing two case-marked pronouns to help them interpret grammatical sentences, both actives and passives, but they were unable to assign agent-patient relationships in any consistent way with ungrammatical pronoun frames. Three-year-olds also used pronoun frames to interpret grammatical active and passive sentences (with either one or two case-marked pronouns) but varied in their interpretation of ungrammatical sentences according to pronoun frame. These results suggest that the role of case-marked pronouns has been underestimated in English language acquisition, and that even very young English children use multiple cues to comprehend transitive sentences.

ACKNOWLEDGEMENTS

This study was made possible by the parents and children who came in to the Max Planck Child Study Centre in Manchester. Many thanks to Dr. Grzegorz Krajewski who assisted in transformations of the data and statistical analyses.

Notes

¹For example, Das Schwein tammt das Zebra (The_neuter pig is tamming the_neuter zebra) which is grammatical but lacks any cues to agent patient assignment based on case marking.

²There is conflicting evidence as to the domination of get- vs. be-passives in British English children's speech (Meints, 2003; Messenger, 2009). It was therefore deemed more reliable to go with just one type.

³We have examined the sentences separately from the one pronoun frame condition, however we found no clear or consistent advantage for sentences with ‘she/he/her/him’ as agent and ‘it’ as the patient over sentences with ‘it’ as the agent and ‘she/he/her/him’ as the patient. We analyze the data in the way that we do for two reasons. Firstly, the motivation for grouping the conditions in the way that we did was on the basis of the relative information content carried by the different pronoun frames (1 case-marked versus 2 case-marked) – the main theoretical motivation of the experiment. Secondly, we argue in the Discussion that animacy effects are neutralized in this experiment, which is perhaps why we do not see any additional performance boost when we do break-apart the 1 case-marked condition and analyze the results separately.