Investigating music as a facilitator of early word learning in autism: Learning words from spoken vs. sung language

Background Although autistic children use the same mechanisms in word learning (WL) as typically developing children1,2,3,4,5,6, they struggle with the generalisation of new words, and there is a lack of research investigating factors that may facilitate learning in children who experience difficulties acquiring language7

. Neuroimaging studies reveal that autistic individuals show reduced activation in language-specific brain regions during speech stimulation, but greater activation during song stimulation8,9. Thus, music could work as a compensator for reduced language abilities9 . To our knowledge, there are no studies investigating if music facilitates WL.

We propose to investigate implicit learning and generalisation of spoken and sung novel words in autism, addressing the following questions: 1. Does music facilitate object-word mappings in autistic children? 2. Does music promote generalisation processes of new words in autistic children? Design We will recruit 30 preschool-aged autistic children, matched with a

neurotypical group on expressive vocabulary. A mixed design will be used, with autism status as between-subjects and WL condition (spoken vs. sung words) as within-subjects independent variables. Number of words learnt receptively and number of generalised novel words will be dependent variables. As social skills and language abilities are closely linked10, and music activates

language-specific brain regions in autistic individuals8,9, children's baseline social responsiveness, language and musical perception will be measured to investigate their role as putative moderators of WL. Children will be tested over three sessions. Baseline measures of potential moderators will be taken during the first session. Tablet-based eye-tracking

measures for sociality and musicality will be developed to minimise task demands. Social responsivity measures will be adapted from an existing app-based assessment for autism11 and music perception measures adapted from existing in- person tasks12. Cognitive and language development will be assessed using

standardised direct and parent report measures13,14 . WL of spoken and sung words will be investigated in two other sessions, adapting an existing WL paradigm2 . Since autistic children benefit from higher iconicity in WL tasks15, materials will include coloured photographs. Warm-up trials will show photographs of familiar objects from MB-CDI word lists14

. Training trials will use nonsense words and novel objects from the NOUN database16. Children will hear and see an animated cartoon character on screen which will name target items (through speaking/singing during the spoken/sung word condition) and ask the child to identify them in retention and generalisation tasks. Eye-tracking technology will be

used to identify children's preferential looking patterns in these tasks. Implications This study will adapt existing assessments to an app-based eye-tracking measure for young children with little or no language which could be used by practitioners to assess social and musical perception skills in the future. Our findings

will provide insight into mechanisms that may facilitate word-object mappings in autistic children experiencing difficulties acquiring language which has implications for speech and language interventions and will help to shape optimal learning environments. The experimental approach using a cartoon character could be

extended to an app-based learning tool to support vocabulary acquisition.