This thesis utilised two distinctive approaches: initially, the categorisation approach was used, followed by a shift to the interactional CA approach. Being aware of the limitations that the categorisation approach has in neglecting potentially meaningful behaviours, the scheme used in original article I was constructed to be more inclusive based on the ideas of Drew et al. (2007). While this allowed for the identification and quantification of gaze behaviours and the comparison of their occurrences across set-tings, more detailed information was lost regarding how gaze was used in interaction and in relation to co-participants’ conduct. This is the case for lot of the research on ASD. Categorisation may produce quantitative information but we might not quite know what it means to have, for instance, more eye contact in one setting than in another. The question, ‘What does gaze do there?’ is left unanswered.
Categorisation and quantification are highly valued in current ASD research based on the biomedical paradigm, as the reported challenges in social interaction are seen to exist ‘within’ individuals with ASD. There has thus been no perceived need to attempt to account for these challenges with ‘external’ explanations, such as examin-ing the actual interactions between individuals with ASD and their social partners.
The reported challenges in social interaction have therefore mainly been studied in controlled laboratory settings. Overall, the biomedical paradigm has different as-sumptions about knowledge production than the social constructionist paradigm.
During the process of preparing the original articles, I encountered countless remarks on ‘lacking’ a control group, the ‘small number’ of participants, ‘missing’ statistics, etc. While such notions are valid in experimentally oriented research, they are less so in a microanalytic, essentially qualitative inquiry that works within a different paradigm. As Schegloff (1993, 115-116, emphases in original) has put it, ‘--- issue is whether there are any distinctive payoffs of quantitative analysis, for the fact that we can do quantitative analysis (if it should turn out to be a fact) does not entail that we should do it’. Instead, it might be beneficial to move away from the ‘what’ question that has dominated the biomedical paradigm, and to engage with the ‘how’ question.
This could also mean shifting the focus from forms of behaviours (e.g. gestures) to functions (i.e. how gestures are used) (see Dickerson et al., 2007), which would allow more focus on interactional processes and contexts, and perhaps deepen the existing understanding of social interaction in ASD. However, the ‘how’ question may not be
relevant for many of those currently engaged with ‘normal scientific’ research. As Kuhn (1970) discusses, new problems might be rejected, as they cannot be addressed with the conceptual and instrumental tools that the paradigm offers. It is thus crucial to continue building dialogue between those conducting ‘normal science’ and those trying out alternative approaches, to encourage engagement with problems that are not defined within the biomedical paradigm. It is through engagement with such new problems that novel discoveries can emerge (Kuhn, 1970).
The isolation approach (Becchio et al., 2010) has also played a key role in the re-search conducted within the biomedical paradigm. The findings of this thesis contrib-ute to the critical voices that have expressed concerns over the valuable information that might get lost when such approaches are applied. Original article III showed how the mere focus on whether a child responds to pointing gestures could distort research results: both the children and their social partners oriented to the way in which the social partners produced pointing gestures for the children to respond.
Thus, excluding social partners’ behaviours or attempting to treat them as neutral seems problematic, both in educational (Dindar et al., 2016b) and assessment settings (Korkiakangas et al., 2016).
The isolation approach has been particularly common in experimental research, such as studies using eye tracking technology. Currently, gaze-related research is mov-ing towards a recognition that ‘certain gaze-related phenomena are interactively con-stituted’ and depend upon ‘participation in social interaction rather than observation’
(Schilbach, 2015, 130). While this shift has not necessarily resulted in a rise in inter-actional research, the interest in live eye tracking in naturalistic settings is increasing (e.g. Falck-Ytter et al., 2015). It should be stressed though that within these live eye tracking studies the ideal of controlling co-participants’ conduct is rarely questioned, and very little attention has been devoted to free flowing interactions between children and their co-participants. As original article IV demonstrated, video-based study of social interaction could complement eye tracking research in a valuable way, without forcing the use of the isolation approach. Such an approach is novel and has been conducted to a limited degree outside the field of ASD (Auer, in press; Hirvenkari et al., 2013; Holler & Kendrick, 2015; Kendrick & Holler, 2017). Categorisation and quantification could be useful tools in combination with the interactional approach, provided that they enable interactionally meaningful discoveries (Schegloff, 1993; see Nishizaka, 2015; Steensig & Heinemann, 2015; Stivers, 2015 for a featured debate).
An increasing number of researchers are engaging with these concerns and con-tributing to the shift towards the interactional approach. That said, it will likely be some time before one is able to talk about an actual paradigm shift: the biomedical paradigm still strongly dominates research on social interaction in ASD, as evidenced by it receiving the most research funding (Deonandan, Liu, Kolisnyk, & Konkle, 2016;
Pellicano et al., 2014).
7.3 imPlications for Practice
The findings of this thesis have implications both for educational and clinical practice.
As original articles II and III showed, adult co-participants have a crucial role in facili-tating interactions involving children with ASD. The CA-based interactional approach can make visible the co-participants’ tacit knowledge that they use in managing their everyday interactions with children with ASD, and enable the examination of which
particular interactional practices work with these children. This information may be useful for educators and efforts could be made to engage them in analysing and reflect-ing on their interactions with the children as part of their professional development.
Similarly, in clinical practice, video recording would provide a useful resource for re-engaging with the ways in which children with ASD produce their responses to test questions. Some, but not all, assessment protocols do encourage the use of video recording during assessment. One such protocol is the ESCS (Mundy et al., 2003) that warns testers to be vigilant and alert when coding eye contact, for example. While such coding would be extremely difficult without a video recording, the examination of that recording can be made significantly more practical by utilising annotating tools (e.g. ELAN) that enable one to review the video using different speeds and to easily search for annotations made earlier.
This thesis also demonstrated that psychometric test results might conflict with findings that are revealed when participants’ (inter)actions are examined in a context-situated manner. This brings into question how professionals define and measure concepts such as interactional competence or intelligence. A good example is Rap-ley’s (2004) interactional work within the social constructionist paradigm that poses criticism of how formal intelligence testing treats people with intellectual disabilities.
He sees these people as having little chance in performing well in such testing situa-tions: ‘If one’s “reliability” as a reporter on one’s own life is to be determined by the asking of bizarre, unoccasioned, questions in artificial encounters with professionals, then it is perhaps not surprising that the disqualification of people with intellectual disabilities as human agents is so readily accomplished’ (Rapley, 2004, 29). Rapley’s notion also applies to people with ASD, and this thesis aligns with his views in en-couraging professionals to be cautious in interpreting psychometric test results and using them to draw conclusions on people’s abilities. These notions have links with what has been described as collaborative/therapeutic or individualised psychological assessment (see Finn, Fischer, & Handler, 2012; Fischer, 2000). These approaches view assessment situations as possibilities for collaboration and intervention – perhaps by discussing testees’ answers to the test questions together - thereby recontextualising the interpretation of psychometric test results. Such approaches thus provide consid-erably more vivid, complex, and accurate pictures of the testees than the summative views provided in Tables 2 (pp. 47) and 3 (pp. 48).
Engaging with how gaze and other multimodal resources are used in assessment interactions, rather than recording their mere occurrence, allows the identification of subtle competencies. Dickerson and Robins (2015) have stressed how this would en-able clinicians to recognise competencies that easily go unnoticed when attention is limited to certain predefined behaviours. Furthermore, they suggest that it could lead to an understanding of the meaning and function of such behaviours, as they would be examined in their context of occurrence rather than in isolation. Such re-examination of taken-for-granted elements of assessment procedures can provide rich information on a child’s abilities and facilitate the development of more reflexive clinical practice (Muskett et al., 2012).
Intervention practices could also be informed by interactional research. Mil-ton (2014) has pointed out that people with ASD often get taught fixed and static social rules in a behaviouristic manner. This does little to help people with ASD to adjust to the complex lived reality where social rules are negotiated (Milton, 2012). Interactional research has the possibility to contribute to designing intervention programmes that include interactional practices that are actually used in everyday interactions. Video
recordings collected during naturalistic and naturally occurring interactions could be utilised in an insightful way with people with ASD to analyse how interactions unfold.
CA has been employed in a similar manner in organisations (Conversation Analytic Role-play Method; Stokoe, 2014), and its use could be extended to increase the social insights of people with and without ASD.
Overall, this thesis suggests that the relationships between individuals with ASD and their social partners should be the core focus of practice. This would allow the use of more collaborative practices, such as with a child’s family, which is essentially a system with interactional and communicational challenges (see Simon, 2016). A child with ASD exists in relation to their social partners, and it is in these contexts, not in isolation, that the challenges in social interaction should be examined and supported.
7.4 limitations and future directions for research
There are some limitations that should be considered. Firstly, there are concerns that relate to the participating children. There were challenges in getting detailed back-ground information, such as psychometric assessment data from the children. The ADOS-2 and the Autism Diagnostic Interview, for example, are considered the ‘golden standard’ and could have been useful. Furthermore, this thesis used no comparative data from TD children. This data would have allowed us to examine whether some of the practices (e.g. repair practices in original article II) were only characteristic of children with ASD, or whether they are also employed by TD children.
In addition, all the participating children were boys, which is common considering the gender ratio of 4:1 in ASD. However, including girls with ASD would have been interesting for examining whether and how their interactions might differ from those of boys. In the future, studies should aim at including more girls with ASD. All the participating children were school-aged which classifies this thesis as a contribution to the small amount of research that has been conducted on this age group.
Secondly, the data for the original articles was recorded in broadly similar tech-nology-enhanced settings, and involved task-related interactions around technology applications between children with ASD or autistic features and their parents, club tutors, educators, and researchers. While this is beneficial in terms of increasing under-standing of task-related interactions, perhaps focussing on interactions taking place at home could be more useful considering the sensitive facilitative work of even the most familiar co-participants. Furthermore, the children mostly interacted with their adult co-participants and much less so with their peers. More research is needed to examine how children with ASD interact with other children with ASD, or indeed those with TD. There is CA-based research that suggests that TD peers might engage less in fa-cilitative work than adults (Rendle-Short, Cobb-Moore, & Danby, 2014), making such interactions potentially more challenging for children with ASD. Also, as technology is increasingly used during free time and in educational settings, it would be fruitful to examine in more detail how technology shapes interactions around its use.
Thirdly, the data cannot be classified as fully naturally occurring due to the fact that the interactions took place during sessions organised as part of two research projects.
Despite this, participants were not encouraged to behave in any particular way, re-sulting in interactions that can be described as naturalistic. While naturally occurring data is the ideal in CA research, the use of certain data collecting methods, such as eye tracking, requires some researcher involvement. However, the data of this thesis
is considerably more naturalistic than that of the majority of research within the bio-medical paradigm, and contributes to calls for the use of more natural environments in ASD research. Studies conducted in natural environments could perhaps make more a significant contribution to the day-to-day lives of individuals with ASD and their families, something that has been called for by stakeholders (Pellicano et al., 2014).
The combination of eye tracking and video-based interactional study of social interaction seems promising and would be a fruitful direction for further research.
The use of mobile eye tracking equipment does not necessitate a technology-enhanced setting, such as the one in this thesis; data can essentially be collected anywhere. Live eye tracking has enormous potential to increase our understanding of gaze-related competencies and challenges in ASD in actual interactions, particularly when com-bined with the interactional approach.
It should be noted that this thesis was framed within a specific discursive psycho-logical reading of social constructionism (Edwards & Potter, 1992; Potter & Edwards, 2013; Rapley, 2004) and built its ideas mostly on the interactional approach. Thus, the focus was on micro level details of interaction rather than on broader matters such as power relations (Gubrium & Holstein, 2008). In line with discursive psychology, this thesis has also taken a critical stance towards the way in which ASD has been con-ceptualised as pathology within an individual, drawing on e.g. cognitive psychology.
The present research was focused on interaction and communication in children with ASD; however, ASD has been also critically approached from different viewpoints fo-cusing, for instance, on diagnostic matters (Timimi & McCabe, 2016b) and knowledge production in ASD studies (Milton, 2014).
Crucially, this thesis can be criticised for not explicitly including the viewpoints of children with ASD or their adult co-participants. Research is moving in such a direc-tion and increasingly participatory approaches are valued (Milton, 2014; Pellicano et al., 2014). As Milton (2014) points out, the involvement of people with ASD in research is a requirement for studies to have full ethical and epistemological integrity. While the CA approach grounds its analysis on participants’ observable conduct and their treat-ment of each other’s conduct, and thus avoids imposing interpretations that would not be relevant to the participants themselves, it is important to consider including people with ASD in the research process. Their insights could reveal more about how they experience social encounters and thus be a fruitful extension to future CA analyses.
7.5 conclusions
At the core of this thesis has been an attempt to encourage a rethinking of some of the ways in which social interaction in ASD has been conceptualised and examined. I have argued that the approaches utilised within the dominant biomedical paradigm have influenced the way in which social interaction in ASD is understood. This view has focused on the difficulties of individuals, without considering the contextual details of interactions in which such difficulties might emerge and the variety of unpredictable ways of interacting that these individuals may employ. Also, the observation that indi-viduals with ASD experience more challenges with TD partners than with those with ASD (Schilbach et al., 2013; see also Milton, 2014) demonstrates that crucial informa-tion can be lost if we only focus on the individuals with a diagnosis and disregard the actions of their TD co-participants as ‘neutral’. To address this problem, I have aimed
to show that social constructionism as a paradigm and, specifically, interactional CA as a methodology, can make a significant contribution to changing how we approach and understand social interactions involving children with ASD.
Shifting the research focus from ‘within individuals’ to ‘in interaction’ provides a means for considering how social interactions actually unfold between individuals with ASD and their social partners, and how they could be facilitated. Such a research focus is challenging to take on under the biomedical paradigm as it is guided by considerably different beliefs, values, and methods than the social constructionist paradigm. Shifting towards the latter and utilising CA would provide scope to move beyond mere challenges when viewing individuals with ASD. As Dickerson and Rob-ins (2015, 76) point out, such a shift might ‘sacrifice something of the security of an apparently standardised tool, but we have the possibility of moving beyond what we expect to find and coming closer to seeing what is actually there in the interaction’.
There is potential for the field of critical autism studies to show that the problems that have pushed the biomedical paradigm towards a ‘crisis’ (in a Kuhnian [1970] sense) can be approached in a new way by taking a critical stance towards taken-for-granted knowledge. This would allow children with ASD to be viewed as active participants that have competencies in initiating, maintaining, and responding to interactions with their social partners.
ReFeRences
Adamson, L. B., Bakeman, R., Deckner, D. F., & Nelson, P. B. (2012). Rating parent-child interactions: Joint engagement, communication dynamics, and shared topics in autism, Down syndrome, and typical development. Journal of Autism and Developmental Disorders, 42, 2622–2635.
Aldaqre, I., Schuwerk, T., Daum, M. M., Sodian, B., & Paulus, M. (2016). Sensitivity to communicative and non-communicative gestures in adolescents and adults with autism spectrum disorder: Saccadic and pupillary responses. Experimental Brain Research, 234, 2515–2527.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Wash-ington, DC: American Psychiatric Association.
American Psychological Association (2010). Publication manual of the American Psychological Association (6th ed.). Washington, DC: American Psychological Association.
Anderson, D. K., Liang, J. W., & Lord, C. (2014). Predicting young adult outcome among more and less cognitively able individuals with autism spectrum disorders. Journal of Child Psychology and Psy-chiatry, 55, 485–494.
Antaki, C., Crompton, R. J., & Walton, C. (2017). How adults with a profound intellectual disability engage others in interaction. Sociology of Health & Illness, 39, 581–598.
Antaki, C., Finlay, W., Walton, C., & Pate, L. (2008). Offering choices to people with intellectual disabilities:
An interactional study. Journal of Intellectual Disability Research, 52, 1165–1175.
Ashwin, C., Hietanen, J. K., & Baron-Cohen, S. (2015). Atypical integration of social cues for orienting to gaze direction in adults with autism. Molecular Autism, 6. doi: 10.1186/2040-2392-6-5
Armstrong, T. (2012). Neurodiversity in the classroom. Strength-based strategies to help students with special needs succeed in school and life. Alexandria, VA: ASCD.
Arnold, L. (2012). Introduction. Autonomy, the Critical Journal of Interdisciplinary Autism Studies, 1. Re-trieved from: http://www.larry-arnold.net/Autonomy/index.php/autonomy/article/view/ED1/
html
Asperger, H. (1944/1991). “Autistic psychopathy” in childhood. Translated by U. Frith. In U. Frith (Ed.), Autism and Asperger syndrome (pp. 37–92). Cambridge, UK: Cambridge University Press.
Auer, P. (in press). Gaze, addressee selection and turn-taking in three-party interaction. In Brone (Ed.), Eye-tracking in interaction. Studies on the role of eye gaze in dialogue. Amsterdam: Benjamins.
Auer, P. (in press). Gaze, addressee selection and turn-taking in three-party interaction. In Brone (Ed.), Eye-tracking in interaction. Studies on the role of eye gaze in dialogue. Amsterdam: Benjamins.