Pragmatic language ability and preferences in autistic and non-autistic children: response content, timing, and social judgements

Social communication difficulties can act as a barrier to peer relationships and social inclusion for some young people, particularly autistic children. Difficulties in this domain, such as engaging in back-and-forth conversation and tailoring behaviour to meet the demands of different social contexts, are part of the diagnostic criteria for autism. However, due to high heterogeneity, there is mixed evidence of how the specific pragmatic language abilities of autistic children may differ from that of their non-autistic peers. There has also been limited exploration of the mechanisms that facilitate communication in both groups. Furthermore, previous research has not investigated how both autistic and non-autistic children's social judgements of others might be influenced by their use of certain conversational behaviours.

Method:

This thesis therefore focuses on children's production of, and social preferences for, specific pragmatic behaviours, concentrating on the domains of referential communication, contingent responding, and conversational response timing. To do this, a range of elicited production paradigms were used, including conversation probes, a virtual Director Task, and a question-answer game. A novel social desirability paradigm was also developed, whereby participants were asked to provide ratings of speakers from scripted, experimentally-manipulated stimuli. Unlike in previous studies that have used snippets of spontaneous conversation, this meant that the specific conversational behaviours that influence children's social judgements could be isolated and identified.

Results:

Regarding children's social impressions, findings from Chapters 2 and 3 suggest that autistic children align with both their non-autistic peers, and neurotypical adults, in indicating that they are less likely to befriend, or enjoy interacting with, a speaker who provides off-topic or delayed conversational responses. Additionally, in Chapter 5, non-autistic children also demonstrated unfavourable social judgements of those with slower speech rates.

Secondly, regarding production skills, findings from Chapters 3 and 4 provide a mixed view of autistic children's pragmatic abilities - whilst they exhibited a reduced ability to provide on-topic conversational responses compared to their non-autistic counterparts, autistic children demonstrated comparable skills in providing appropriately-informative referential statements, and timely responses.

Finally, regarding factors affecting children's pragmatic abilities, findings from Chapters 4 and 5 identified the possible impact of increased cognitive load on both response content and timing. That is, in Chapter 4, increasing the cognitive demands of the Director Task by adding more competing objects resulted in a smaller proportion of appropriately-informative responses. Also, in Chapter 5, when the speech rate of pre-recorded questions was experimentally manipulated, participants responded earlier to slow questions than fast questions, potentially reflecting the higher language processing demands associated with faster speech.

Discussion:

Overall, findings from this thesis contribute to our understanding of group differences in the pragmatic abilities of autistic and non-autistic children, as well as highlighting the potential impact of cognitive load on these skills. This thesis also emphasises the social importance of children's pragmatic language abilities. That is, it highlights how a child's tendency to exhibit unconventional conversational behaviours - off-topic, delayed, or slow responding - could result in negative social impressions from peers, which can act as a barrier to friendships and inclusion.