Why young children fail to understand ‘before’ and ‘after’

The goals of the thesis were to identify the development of 3- to 7-year-old children’s comprehension and production of two contrasting temporal connectives - before and after - that signal the order of events in two-clause sentences, and to establish the reasons for difficulties with these linguistic devices. Chapter 1 reviews the literature that is considered relevant to the experimental work. In the experimental work (Chapters 2 to 4), children’s comprehension and production of two-clause sentences containing before and after was examined in separate groups of children aged 3 to 7 years. The sentence structures differed in their memory and also language demands. Independent measures of memory and language were related to performance. The design enabled a contrast of traditional memory capacity accounts (e.g., Just & Carpenter, 1992) versus more recent language-based accounts (e.g., Van Dyke, Johns, & Kukona, 2014) of why working memory explains variance in the processing of complex sentences. A capacity account predicts a direct relation between memory and sentence processing: specifically, that some sentence structures are more difficult to process than others because they require more information to be held in working memory than others. Alternatively, a language-based account proposes an indirect relation between memory and sentence processing, such that good language skills modulate the influence of memory on sentence processing, by influencing the accurate representation of information in verbal working memory. Experiment 1 (Chapter 2) was a touch-screen comprehension paradigm. Children listened to two-clause sentences linked by a temporal connective, before or after, while viewing animations of the actions in each clause. After each sentence, they were asked to select the event that happened first to assess their understanding of the temporal connective. The pattern of results suggested that the memory demands of specific sentence structures limited children’s comprehension of sentences containing temporal connectives, supporting a memory capacity account. Experiment 2 (Chapter 3) further investigated comprehension of these sentences focusing on how memory and language influence the ease of processing. Children were trained to make speeded responses to the sentence structures investigated in Experiment 1. The findings support Experiment 1: memory capacity best predicted comprehension of these sentence structures. Experiments 3 and 4 (Chapter 4) examined production of the same sentence types. In two experiments (elicited production with blocked conditions, and sentence repetition), separate groups of children viewed an animated sequence of two actions, and were asked to describe the order of events. Instructions and practice trials were used to model the target sentence structures. In contrast to the comprehension experiments (Experiments 1 and 2), this work showed that children’s individual differences in the production of two-clause sentences linked by before or after were related to variability in language skills, rather than poor memory capacity. In Chapter 5, I conclude that Experiments 1-4 reveal a differential influence of working memory and language on children’s comprehension and production of two-clause sentences containing before and after. I argue that the existing theoretical accounts of the influence of memory and language on sentence processing (e.g., Just & Carpenter, 1992; Van Dyke et al., 2014) require much more detailed investigation within the sentence structures examined here, and across other complex sentences that are also considered to differ in their memory and language demands. I present several suggestions as to how this might be accomplished in future work