Autism Technology Research has a Problem: They’re Designing for Neurotypical People
This article comes with a content note on violent ableism.
Technology research follows a dangerous mantra claiming that with the preferences for sameness and predictability, autistic people are uniquely adept with technological environments and supposedly prefer them over human contact. The short-sighted notion ignoring the multi-facetted ways in which autistic people interact with the world in general and others specifically keeps being perpetuated for the sake of argument towards funding bodies and within research papers. And technology research around autism, particularly in the context of autistic children, is exploding. Hence, it makes sense to look into the role autistic people actually play in this research, or so I thought.
As part of my PhD research, I conducted a large scale literature review looking at technologies addressing autistic children aged 6–10 (mostly to get some kind of handle over the size of the corpus). By now, the associated paper (co-authored with Chris Frauenberger, Os Keyes and Geraldine Fitzpatrick) has been published; this post comprises a summary of our main findings and conclusions.
I started out with 2083 papers and after iterative refinements of focus, ended up with 185 that I’ve read closely. My interest was in finding out what all these technologies were essentially for and what kind of purpose they followed. In the end, I established six categories: behaviour analysis, assistive technologies, education, social skills, therapy and well-being. In the paper, we then exercise what each of them entails.
In the context of technologies conducting behaviour analysis, we found that “autistic children are secondary users (…). The technologies act here as mediators for the adults’ interpretation of the children’s behaviour and facilitate further decision making that impacts the lives of autistic children” (p.15). The subset of assistive technologies conceptualises autistic children as primary users. Though, this comes at a catch, for example, when it comes to communication devices: “The communication is inherently limited to aspects that designers of these technologies have implemented. As autistic children are sometimes included in the design of these technologies, this could be a non-issue; however, for autistic children who use these technologies in their everyday life, the range of assistance is limited to the needs of an outside world or what that world assesses as a need of the child” (p.16). This is similarily the case for technologies in the context of education (learning materials tend to be (ironically!) restricted to narrow interests without supporting self-guided exploration) and social skills training (single-sidedly focused on ‘teaching’ normative skills). The purpose of therapy doesn’t quite allow for this distinction between primary and secondary users; it’s also the one that almost broke me when I head to read through all the papers. They emphasise the fundamental intent to change an “autistic child at such great expense that it presupposes that the child is implicitly conceptualised as an entity that is in need of ‘correction’ and ‘improvement’, without including a notion of acceptance of different ways of being in the world” (p.19). Finally, within the well-being category, we only found highly individualised prototypes and large installations, indicating that none of these are easily available to a larger group of children without depending on the buy-in from adults.
Across these purposes, we can see how, in their entirety, they paint a bleak picture. Autistic children are persistently defined along medicalised assumptions of deficits and disorders and these definitions carry over into technological developments that while claiming to augment, help, assist and support end up restricting, limiting, ignoring and othering autistic modes of communicating, perceiving, socialising and being.
Our conclusion is that current autism technology research has an enormous conceptual problem: It is not actually intended for autistic people. Instead, it “targets the normative concepts and desires of their mostly neurotypical social environment. (…) The majority of technologies and the research around them manifest a perspective that conceptualises disabled people as passive and without agency. Autistic children in combination with these tech- nologies conceptually establish a socio-technical construct that an allistic environment interacts with. As the technologies are designed, introduced and facilitated by allistic stakeholders, access and interaction are equally restricted and regulated” (p.23). We urge researchers in the field to do better, to take their responsibilities seriously and to reflect on the larger implications of their research. To attend to autistic people’s needs, desires, wishes and preferences, to admit that one can be wrong, to acknowledge difference in appreciation and to be honest about their work. Allistic researchers designing technologies in an autism context, could think of them as being of service, practice humility, carefully deprioritise neurotypical positions in their research and only then might they claim to design technologies for autistic people.
