Thursday, December 2, 2021

The Girl Was Watered by the Flower: Effects of Working Memory Loads on Syntactic Production in Young Children

Adams, & Cowan, N. (2021). The Girl Was Watered by the Flower: Effects of Working Memory Loads on Syntactic Production in Young Children. Journal of Cognition and Development, 22(1), 125–148. https://doi.org/10.1080/15248372.2020.1844710

Working memory is the ability to maintain information in mind while manipulating that (or other) material in some way. Working memory is important for language production and comprehension tasks such as sentence repetition. Sentence repetition draws on language and working memory skills in order to hear and understand the sentence, retain the meaning of the sentence in mind, and then formulate and produce the sentence for recall. Working memory resources may be needed differently for different types of sentences, however. For instance, after hearing a passive sentence (the flower was watered by the girl), less working memory resources might be needed to use the more familiar active form (the flower was watered by the girl). On the other hand, it could be that verbatim repetition of passive sentences is easier than transforming the passive sentence into an active form. The aim of the current study was to understand how working memory loads affect recall of passive sentences.

Across two studies, children ages 4-6 completed a sentence repetition task with and without a memory load. An auditory-verbal memory load was imposed by having children store a list of spoken digits in mind. A visuo-spatial memory load was imposed by having children store a series of spatial locations in mind. Experiment 1 unfolded as follows: (1) children listened to passive sentences, (2) retained the memory load, (3) described the same pictures as shown in (1) (children were not explicitly told to use a passive voice), and (4) recalled the memory load by saying the digits or drawing the location. In Experiment 2, children were explicitly instructed to use the passive voice in step (3) of the procedure.

The results of Experiment 1 suggested that children were likely to use passive sentences with and without a memory load. In Experiment 2, children were more likely to use the passive voice with a visual-spatial load and active sentences without a memory load. Because passive sentences were retained even with reduced working memory resources, the authors suggested that children were simply repeating sentences verbatim. Additional support for simple repetitions came from the types of errors children made. Although children were using the passive voice, they made semantic errors such as switching the roles of the nouns (the girl was watered by the flower).

The results suggest that children can repeat sentences verbatim without understanding them, whereas working memory resources are required to make more semantically accurate responses. Given that immediate sentence recall tasks are used as part of language assessment in speech-language pathology, it is important for clinicians to understand the influence of language and memory on performance. As well, it is important to consider going beyond scoring number of errors to potentially understanding the types of errors being made.


Theresa Pham completed the combined SLP MClSc/PhD program under the supervision of Dr. Lisa Archibald.