“Okay, so Anyway”: The Influence of Discourse Markers on Pragmatic Language in the Broad Autism Phenotype

Background: Pragmatic language is universally impaired in individuals with ASD, and sub-clinical pragmatic language differences have been observed in a subgroup of first-degree relatives. Believed to constitute an endophenotype indicative of genetic risk to ASD, pragmatic language differences among relatives have been described as part of the broad autism phenotype (BAP). Prior investigations of pragmatic language in unaffected relatives revealed global differences in contingency and content in both narrative and conversational tasks (Landa et al., 1991, 1992; Losh et al., 2008, 2012).  This study attempted to build on this work by applying a fine-grained discourse analysis to transcripts of conversational interactions from parents of individuals with ASD. In particular, we examined the use of conversational markers essential for cohering messages and maintaining the conversational floor, as well as the relevance and repetition of thematic content, which influences conversational reciprocity.   

Objectives: To examine the patterns of discourse marker use and thematic content of conversational samples from parents of individuals with ASD and age- and IQ-matched controls.   

Methods: Participants included 30 parents of individuals with ASD and 36 parents of typically-developing controls. A detailed coding scheme (adapted from Martin et al., 2012) was used to assess discourse marker use, thematic content, and topic repetition during a semi-structured conversation. Discourse markers included Back-Channeling (i.e., listener responses, ‘okay’, ‘yeah’) and Filled Pauses (e.g., ‘uh’, ‘um’). Thematic content was evaluated by how related a topic was to ongoing discourse (i.e., Contingent vs Noncontingent). Topic repetition was derived by the proportion of topics that were spontaneously re-introduced by participants >2 times. BAP status (positive + or negative -) was determined using the Modified Personality Assessment Scale (MPAS; Tyrer, 1988) and global pragmatic language was assessed using the Pragmatic Rating Scale (PRS; Landa, 1992).  

Results: Family diagnostic groups differed in their use of discourse markers. Specifically, the ASD parent group overall used Filled Pauses less than Control parents F (1, 79) =6.44, ps<.05. When broken down by BAP status, however, the BAP (+) parents used Back-Channeling significantly less than control parents F (2, 57) =4.45, (p<.01) and BAP (-) parents (p<.02). BAP (+) parents also demonstrated increased discussion of Noncontingent topics F (2, 66) =6.95, ps<.01 relative to both groups. Greater discussion of Noncontingent topics was significantly related to a pragmatic language domain associated with talkativeness and overly detailed language (r =.63) in BAP (+) parents. Higher rates of repetitive topic introduction were observed in the BAP (+) group F (2, 67) =2.40, ps<.05.   

Conclusions: Differences in discourse marker use and thematic content in conversation were associated with subclinical features of genetic risk in parents of individuals with ASD. In particular, BAP (+) parents demonstrated word use patterns indicative of reduced reciprocity. Such discourse patterns may significantly, albeit subtly, influence communication styles relevant to interpersonal relationships. These findings underscore the subtle manifestation of pragmatic language differences as a possible endophenotype of ASD.