Variability in Child Speech (VariCS)

This project will explore how speech changes as children grow. Learning to speak is one of the biggest developmental tasks, and children vary considerably in how quickly they master all aspects of this complex process. Irrespective of what language or dialect a person uses, producing speech first involves controlling the breath as it leaves our lungs (respiration). We control this air as it passes through our voice box (phonation) and use the tongue, jaw and lips to form specific speech sounds (articulation) that come from the sound waves in our throat, mouth and nose (resonance). Together these aspects are known as the four 'speech subsystems'. Whilst we know a lot about infant speech, as with many other areas of child development, fine-tuning the workings of these subsystems takes place well into adolescence. Currently we know very little about how primary-school aged children coordinate their use of the four speech subsystems, and how much children vary in terms of how and when their speech production patterns mature.

However, this variation is important for us to better understand the development of speech in children with speech disorders. Some children produce unclear speech or have a weak sounding voice. This may be the result of facial differences such as cleft palate, or damage to the brain that control the muscles needed for producing speech, which is the case in children with cerebral palsy. These physical problems can affect how well the different speech subsystems work together. For Speech and Language Therapists (SLTs) to be able to provide adequate treatment we need to know to what extent children with speech disorders differ from typically developing children. This is why we need better knowledge on the role of variation in typical child speech as children grow.

To achieve this, the Variation in Child Speech (VariCS) project will use a longitudinal approach. This means we will repeatedly collect information on the four speech subsystems in 500 primary-school children over two years, obtaining data for each age group from 4 years to 11 years. We will ask children to tell a story, repeat sentences and words, and make long vowel sounds. The data will be collected from children located in the central belt of Scotland, but the nature of the speech tasks and our analyses means that the results will be applicable to typical and disordered speech for many dialects and languages. We will record children's speech using an app to make sure all children do the same tasks and have fun when recording their speech. Based on the recordings, we will gain essential information about the functioning of each speech subsystem, i.e. respiration, phonation, resonance and articulation. For instance, producing a long vowel sound will provide information on a child's lung capacity, which helps us to examine their respiration. We will analyse the speech to find out: i) how speech measures obtained for each subsystem change over time, and ii) whether changes in one subsystem can predict changes in another.

The project will provide the first comprehensive information - for all four subsystems - for the developmental pathways of speech production in children aged 4 to 11. At the end of the project, we will have i) comprehensive knowledge on child speech development and its variability over time, allowing us to compare disordered speech to typical speech, ii) information about how much typical speech varies between children, which we will collate in an interactive web resource for SLTs, and iii) a large speech data set that will be made available for further research after the project.