A key recommendation from Jaswal & Akhtar (J&A) concerning how to reform the science of autism – how to make it more humane, accurate, and useful – is to take seriously autistic testimony. It is not uncommon to incorporate autistic testimony into autism science by using examples drawn from such testimony as illustrations. But the authors go beyond recommending a merely illustrative role for autistic testimony. They are calling for more substantial ways to take it seriously, ways that would shape how the science of autism is done (e.g., what scientists investigate and how they investigate what they investigate, including what questions they ask and what assumptions they make) and help identify alternatives to taking a deficit view of autistic differences, which can impede scientific progress in our understanding of autism (Dinishak Reference Dinishak2016). Below I pose two questions to clarify and develop J&A's proposal and conclude with a general lesson about the value of autistic testimony that we should glean from their arguments.
First, in making the case for giving autistic testimony a more substantial role in autism science, J&A characterize autistic people as an “essential … source of insight about autism” (sect. 1). But in what sense precisely are autists an essential source of insight about autism? Is the idea that the only route to acquiring such insight is through autists’ self-reports, such that insight cannot be gotten any other way? Does it mean that the insight gleaned from taking autistic testimony seriously has special value? If that is correct, what kind of value is this? Is such insight necessary input in the science of autism, for example, when assessing the relevance and explanatory power of hypotheses concerning autists’ atypical behaviors? Our answers to these questions have implications for theory choice in autism science – in particular, whether taking autistic testimony seriously should be thought of as merely a recommendation that scientists could rationally and ethically choose to ignore or as a requirement for the science of autism reform called for by the authors.
My second, related question concerns precisely how to understand the particular roles autistic testimony should play in autism science. The following are among the roles for autistic testimony recommended by the authors:
1. To help identify and test assumptions about autism and autistics
2. To help identify unconventional and idiosyncratic ways autistics show their desire to engage with others or express social interest
3. To help assess existing explanations for atypical behaviors and identify alternative explanations for those behaviors
I will focus on the third one.
Toward the end of the target article, J&A write that they hope to have made a case for “the importance of taking seriously the phenomenological experiences” (sect. 6) of autists. This statement provides an intriguing lead on how to understand the third point. It suggests that autistic testimony contains valuable phenomenological data that should be taken seriously by autism researchers. One way to use autistic testimony to assess explanations of autists’ atypical behaviors, then, is to use the phenomenological data obtained from such testimony as a control on theorizing about autists’ social behavior; one checks that an explanatory hypothesis is consistent with phenomenological data obtained from autists’ self-reports of how they experience social interaction.
To clarify and develop this lead, we need to consider the different ways of reading should in “phenomenological data should be taken seriously by autism researchers” (see preceding paragraph), given that different readings may have different methodological implications. For example, the authors note approvingly (sect. 6) that Hadjikhani et al.’s (Reference Hadjikhani, Johnels, Zürcher, Lassalle, Guillon, Hippolyte, Billstedt, Ward, Lemonnier and Gillberg2017) neurobiological findings are consistent with autistic self-reports about eye contact feeling uncomfortable, as Hadjikhani et al. themselves note. This suggests that, at the very least, it is an attractive feature of a scientific account of autists’ social behavior if it is consistent with phenomenological data obtained from autists’ self-reports of social experience. Stronger construals of “should” are that consistency with the phenomenological data is preferable or even required for acceptance of an explanation, in this case regarding autists’ atypical behaviors.
A different way to develop the third point, also suggested by J&A's analysis, is to understand autistic testimony as an explanandum, as evidence that a theory of autism ought to explain: “One readily available source of evidence that the social motivation perspective does not explain is the testimony of many autistic people who claim to be socially motivated” (sect. 3).
The particular nature of the constraint (e.g., consistency with versus ability to explain data obtained from autistic testimony) and the constraint's strength matter for determining the appropriate action when an explanatory hypothesis under assessment fails to meet the constraint. For example, does failure to meet the constraint constitute decisive grounds for rejecting a candidate explanation?
As we can see, J&A's discussions generate many important and provocative questions concerning how best to pursue their proposal for taking seriously autistic testimony in autism science. Their general argument – that it is vital, in terms of improving autism science and more generally, to explore and cultivate a rich appreciation of the ways our knowledge of autism is dependent on autists themselves and on what autists tell us about what it is like to be autistic – is persuasive and epistemically and ethically significant, regardless of how we answer the particular questions raised above. Through their own use of autistic testimony, J&A demonstrate that it is an epistemological resource. Through their critique of the social motivation perspective they show that there is not just epistemological value in taking autistic testimony seriously, there is also moral value in doing so. All in all, taking autistic testimony seriously can help make the science better at finding out about autism (e.g., by exposing biases and unwarranted assumptions) and more ethically responsible (e.g., by addressing a neglect of alternative explanations for autists’ atypical behaviors communicated by autists themselves and by helping to address existing and prevent further epistemic injustice – harm or wrongdoing to autists in their capacity as knowers (Fricker Reference Fricker2007)). Ethical justice and epistemic justice are interdependent pursuits (Grasswick Reference Grasswick, Kidd, Medina and Pohlhaus2017) in the science of autism.
A key recommendation from Jaswal & Akhtar (J&A) concerning how to reform the science of autism – how to make it more humane, accurate, and useful – is to take seriously autistic testimony. It is not uncommon to incorporate autistic testimony into autism science by using examples drawn from such testimony as illustrations. But the authors go beyond recommending a merely illustrative role for autistic testimony. They are calling for more substantial ways to take it seriously, ways that would shape how the science of autism is done (e.g., what scientists investigate and how they investigate what they investigate, including what questions they ask and what assumptions they make) and help identify alternatives to taking a deficit view of autistic differences, which can impede scientific progress in our understanding of autism (Dinishak Reference Dinishak2016). Below I pose two questions to clarify and develop J&A's proposal and conclude with a general lesson about the value of autistic testimony that we should glean from their arguments.
First, in making the case for giving autistic testimony a more substantial role in autism science, J&A characterize autistic people as an “essential … source of insight about autism” (sect. 1). But in what sense precisely are autists an essential source of insight about autism? Is the idea that the only route to acquiring such insight is through autists’ self-reports, such that insight cannot be gotten any other way? Does it mean that the insight gleaned from taking autistic testimony seriously has special value? If that is correct, what kind of value is this? Is such insight necessary input in the science of autism, for example, when assessing the relevance and explanatory power of hypotheses concerning autists’ atypical behaviors? Our answers to these questions have implications for theory choice in autism science – in particular, whether taking autistic testimony seriously should be thought of as merely a recommendation that scientists could rationally and ethically choose to ignore or as a requirement for the science of autism reform called for by the authors.
My second, related question concerns precisely how to understand the particular roles autistic testimony should play in autism science. The following are among the roles for autistic testimony recommended by the authors:
1. To help identify and test assumptions about autism and autistics
2. To help identify unconventional and idiosyncratic ways autistics show their desire to engage with others or express social interest
3. To help assess existing explanations for atypical behaviors and identify alternative explanations for those behaviors
I will focus on the third one.
Toward the end of the target article, J&A write that they hope to have made a case for “the importance of taking seriously the phenomenological experiences” (sect. 6) of autists. This statement provides an intriguing lead on how to understand the third point. It suggests that autistic testimony contains valuable phenomenological data that should be taken seriously by autism researchers. One way to use autistic testimony to assess explanations of autists’ atypical behaviors, then, is to use the phenomenological data obtained from such testimony as a control on theorizing about autists’ social behavior; one checks that an explanatory hypothesis is consistent with phenomenological data obtained from autists’ self-reports of how they experience social interaction.
To clarify and develop this lead, we need to consider the different ways of reading should in “phenomenological data should be taken seriously by autism researchers” (see preceding paragraph), given that different readings may have different methodological implications. For example, the authors note approvingly (sect. 6) that Hadjikhani et al.’s (Reference Hadjikhani, Johnels, Zürcher, Lassalle, Guillon, Hippolyte, Billstedt, Ward, Lemonnier and Gillberg2017) neurobiological findings are consistent with autistic self-reports about eye contact feeling uncomfortable, as Hadjikhani et al. themselves note. This suggests that, at the very least, it is an attractive feature of a scientific account of autists’ social behavior if it is consistent with phenomenological data obtained from autists’ self-reports of social experience. Stronger construals of “should” are that consistency with the phenomenological data is preferable or even required for acceptance of an explanation, in this case regarding autists’ atypical behaviors.
A different way to develop the third point, also suggested by J&A's analysis, is to understand autistic testimony as an explanandum, as evidence that a theory of autism ought to explain: “One readily available source of evidence that the social motivation perspective does not explain is the testimony of many autistic people who claim to be socially motivated” (sect. 3).
The particular nature of the constraint (e.g., consistency with versus ability to explain data obtained from autistic testimony) and the constraint's strength matter for determining the appropriate action when an explanatory hypothesis under assessment fails to meet the constraint. For example, does failure to meet the constraint constitute decisive grounds for rejecting a candidate explanation?
As we can see, J&A's discussions generate many important and provocative questions concerning how best to pursue their proposal for taking seriously autistic testimony in autism science. Their general argument – that it is vital, in terms of improving autism science and more generally, to explore and cultivate a rich appreciation of the ways our knowledge of autism is dependent on autists themselves and on what autists tell us about what it is like to be autistic – is persuasive and epistemically and ethically significant, regardless of how we answer the particular questions raised above. Through their own use of autistic testimony, J&A demonstrate that it is an epistemological resource. Through their critique of the social motivation perspective they show that there is not just epistemological value in taking autistic testimony seriously, there is also moral value in doing so. All in all, taking autistic testimony seriously can help make the science better at finding out about autism (e.g., by exposing biases and unwarranted assumptions) and more ethically responsible (e.g., by addressing a neglect of alternative explanations for autists’ atypical behaviors communicated by autists themselves and by helping to address existing and prevent further epistemic injustice – harm or wrongdoing to autists in their capacity as knowers (Fricker Reference Fricker2007)). Ethical justice and epistemic justice are interdependent pursuits (Grasswick Reference Grasswick, Kidd, Medina and Pohlhaus2017) in the science of autism.