Neurodiversity, Advocacy, Anti-Therapy
Chapman, R. & Bovell, V. Forthcoming. Handbook of Autism and Pervasive Developmental
Disorder. Sturmey, P. & Matson, J. (eds.) Springer.
This chapter provides an overview of the origins and theoretical stance of the neurodiversity
movement, and then explores the implications in terms of what, if any, treatments for
autism are appropriate. We outline some of the key arguments and critiques made by
neurodiversity proponents, and as a case study we focus on the controversy around Applied
Behaviour Analysis (ABA) – the most widely used approach in North America. We discuss
four distinct issues: efficacy, normalisation/social control, trauma/wellbeing, and autonomy.
After examining each, we argue that a significant element of contemporary ABA practice
requires abolition or reform, and suggest that whether or not ABA can be rehabilitated in
the eyes of its critics remains to be seen. We further suggest that the debate around ABA
has implications for the practice and research around all forms of autism therapy. In
particular, we propose that more discussion between neurotypical and neurodivergent
people needs to take place to establish what goals and outcomes are beneficial for whom,
and for what purpose, and in what context.
Keywords: Neurodiversity; Applied Behaviour Analysis; ethics; bioethics; autism treatment
At one time, autism advocacy was largely associated with parents and doctors advocating
for greater autism awareness, increased funding for services, and more research into
treatment and therapy (Feinstein, 2010). Roughly from the early 1990s onwards, autistics
began to gain more recognition advocating for themselves. In contrast to earlier advocacy,
they increasingly pushed against the idea that autism was necessarily a tragedy (Sinclair,
1993) and became more interested in autism acceptance. Out of this grew the
neurodiversity movement, a social justice movement that began among autistic self-
advocates but which has since become associated with a paradigm shift regarding cognitive
and developmental disability more generally (Chapman, 2019; Singer, 1999; Walker, 2012).
Prior to the rise of the neurodiversity movement, autism advocacy had tended to assume
that autism should be viewed as a medical condition – be this psychiatric or neurological. In
contrast to this medical paradigm, neurodiversity proponents were influenced by the
broader disability rights movement and their politicisation of disability. Because of this, they
viewed autistic disablement and distress more as a broader social and political issue
(Chapman, 2019a; Singer, 1999;). The arguments made by neurodiversity proponents are
varied, and it is important to note that the movement has mostly grown online in blogs and
social media more rapidly than through scholarly journals. While there have been influential
figures, the movement has nothing like a textbook or leader at the time of writing. Still, the
movement’s politics are underpinned by some core philosophical and ethical arguments.
Perhaps most centrally, it is relatively common for autistic neurodiversity proponents to
challenge the depiction of psychological disabilities as inherently pathological. This is usually
done by drawing on constructivist models of disability that explain disablement as socially
and culturally produced. At the individual level, many autistic individuals have come to see
autism as an inherent part of their selfhood, and even a political identity, rather than a
disorder to be treated or cured (Chapman, 2016; Sinclair, 1993). Hence autistic people tend
to use identity first language (i.e. “I am an autistic person”) rather than person first language
(“I am a person with autism”). We shall adhere to this preference here. In turn, at the group
level, neurodiversity proponents reframe autism and other similar disabilities as
“neurominorities” rather than as mental disorders (Walker, 2012). On the neurodiversity
perspective, society has been mainly built for and by “neurotypicals”, in such a way that its
dominant notion of “normality” inevitably disadvantages, devalues, and disables
neurodivergent individuals and neurominority groups. Autistics as a neurominority thus
form a marginalised and oppressed group, somewhat akin to minority ethnicities or sexual
To the extent that this reframing is feasible, it raises significant ethical issues when it comes
to therapeutic or clinical practice, not to mention public policy. On the one hand, if it is true
that autism is a medical pathology that causes significant harm to autistic individuals and
others around them, then it seems right to try and treat, cure, or prevent it. Arguably, to not
do so would be wrong, just as it would be wrong to refrain from trying to treat cancer. This,
implicitly at least, has been the orthodox view of autism since it was first described in the
1940s. By contrast, if the neurodiversity perspective is convincing, then such attempts may
be just as bad as trying to “cure” being gay. That is, they would be deeply unethical and
based on a catastrophically misguided undue pathologisation of a minority way of being.
Indeed, for many neurodiversity proponents, the proliferation of genetic investigations,
often explicitly justified as raising the possibility of preventing autism, is reminiscent of
eugenics, and must be resisted at all costs.
Here we will outline some of the key arguments and critiques made by neurodiversity
proponents before turning to the issue of treating or curing autism. Our more specific focus
here is the controversy around Applied Behaviour Analysis (ABA). This method is the most
widely used for autism in North America and is often described as the gold standard, yet is
hotly contested by many autistics and in particular neurodiversity proponents, who see it as
akin to conversion therapy. After reviewing the existing evidence of both the harms and
benefits associated with ABA, we pose a challenge to the ABA industry with respect to
autism: Either concede that the wholesale application of ABA to autism lacks moral
permissibility, or provide evidence that there is at the very least a subsection of the autistic
population who benefit, bearing in mind the associated risks we discuss below. More
broadly, we challenge all approaches to autism intervention and research to consider the
notion of ‘benefit’ in terms of well-being and autonomy, as defined in terms consistent with
perspective of autistic people and those closest to them, as well as being foundational
principles in bioethics.
The Medical and Social Models
Traditionally autism, like other disabilities, has been viewed through what is sometimes
called the ‘medical model’ of disability. Although there are many variations of this model,
what they share is that they frame disability chiefly as a problem that primarily stems from
an individual’s impairments. When it comes to autism and other neurominorities, the
explicit claim of the medical model (e.g. in the DSM) is that the behavioural problems must
stem from the dysfunction within a person, rather than from a mere clash with external
conditions such as societal norms (American Psychiatric Association [APA], 2013). While this
model does take into account how external issues, such as stigma or poverty, can also affect
the lives of disabled people, it is distinguished by its framing of disability as being primarily
caused by individual biological or psychological impairments or dysfunctions.
On this model, autism is viewed as a spectrum disability identified in light of social and
communication issues coupled with highly restricted and repetitive habits, actions, and
routines and sensory sensitivities (APA, 2013). These are seen as harmful deficits, stemming
from underlying dysfunction at the psychological and biological levels (Wakefield, 1992).
Some theories have associated autism with impairments in empathy or cognition, which are
both taken to have a largely biological basis and in turn to cause autistic disablement. In this
view, autism is something that an individual has, much as one can have cancer. A cancer
patient is not described as “cancerous”, because cancer is viewed as an unwanted addition
to the person rather than part of them. Similarly, by saying someone “has autism”, the
implication is that autism is an add-on, and a problematic one, rather than being regarded
as merely a different way of being.
While the medical model is still dominant, in the 1970s disabled self-advocates began to
challenge this by developing the ‘social model’ of disability (Oliver, 1990). This model also
has many variations, but the key difference is that it distinguishes between “disability” and
“impairment”, and frames disability as often, or in large part, being caused by societal
unwillingness to accommodate impaired individuals. By distinguishing between impairment
and disability, disabled self-advocates were able to frame the related distress that disabled
individuals experience as being primarily caused by broader social pathology, rather than by
The social model was initially mainly applied to physical disability, leading to important
social reforms in countries where it has been implemented (Oliver, 1990). For instance, the
implementation of ramps and disabled parking spaces stemmed from shifting the focus
from the individual to the environment. However, neurodiversity proponents have applied
the social model to autism, highlighting how attitudes, practices, and structures contribute
to autistic disablement (Chapman, 2019a). For instance, when it comes to physical
structures, it may be that workspaces are designed in such a way that they are disabling to
autistic individuals with heightened sensory sensitivity (Booth, 2016). And when it comes to
practices, how we organise and teach in classrooms is designed for neurotypical processing
styles and may exclude autistics, especially those with further cognitive disabilities
(Goodley, 2018). Social norms and attitudes are also relevant. Mitz (2008) argued that “the
identification of autism with self-loss, otherness and dehumanisation pervades the
dominant, medicalised discourse of autism” (p. 21). In line with this, Cage et al. (2018)
found that autistic people were associated with higher levels of dehumanisation than
Building on the theorising of autistic self-advocates, the social model has also been applied
to autistic wellbeing. While autistic well-being is, on average, lower than the average,
neurodiversity proponents have argued in detail what we can understand this as stemming
in significant part from external factors (Robertson, 2010). In line with this, Renty and
Royers (2006) found that well-being was predicted by how well supported autistic
individuals felt rather than by how impaired they were taken to be. More recently, Frost and
Botha (2018) studied autism using the minority stress model, which had previously been
applied to sexual and ethnic minorities in order to explain mental and physical health issues
as arising from disparities between minority and majority groups. In the autistic population,
they found that “Minority stressors such as victimization and discrimination, everyday
discrimination, expectation of rejection, outness, internalized stigma, and physical
concealment of autism consistently predicted diminished well-being and heightened
psychological distress” (2018, p. 27).
To be clear, in making the shift to the social model, neurodiversity proponents do not claim
that that there are never any impairment effects associated with autistic cognitive
(Ne’eman, 2010; Chapman, 2020). Kapp et al. (2013) found that while most autistic
neurodiversity proponents were against trying to cure autism as such, they did not find
ameliorating specific more disabling impairments to be a problem. Walker (2012) has
argued that neurodiversity proponents are more concerned with cognitive variations that
underlie different forms of selfhood. Hence, something pervasive such as the sensory
differences associated with autism should not be “treated” since they are constitutive of a
“way of being” (Sinclair, 1993). By contrast, epilepsy, gastrointestinal issues, and other
problems associated with autism (co-occurring conditions, or “specifiers” within the DSM)
should be considered genuine medical problems, and hence the subject of medical
Autism Knowledge Production
Importantly, neurodiversity proponents do not just disagree with the idea that autistic
impairments are the key cause of autistic disablement and distress. They also challenge how
autism is characterised and defined. On the medical model, framing autism has typically
been defined in light of deficits. At the cognitive levels, this includes empathy deficits, weak
context processing, and sensory processing impairments. At the behavioural level, diagnosis
is only framed in terms of problems. Neurodiversity proponents have criticised the medical
model as leading to a depiction of autism based too heavily on deficits rather than benefits
(Dinishak, 2016). They have also challenged the “tragedy” narrative they associated with the
medical-deficit view, which they take to be misleading and harmful for the autistic
population (Sinclair, 1993). Recent advances in neurodiversity theory also challenge the
notion of “impairment” due to issues in its conceptualisation (Chapman, 2020).
Here the critique of knowledge production in the autism industry is key. According to
Walker, the fundamental axioms and assumptions underlying what he terms the “pathology
paradigm” (2012) are normatively and ideologically laden in such a way that will be
reflected in the results of autism science. By this he means that if a scientist’s starting-point
is that difference is a sign of dysfunction – rather than being simply a different form of
functioning – then any findings that demonstrate difference will serve as proof of a
problem, rather than merely an indication of divergence. In line with this, autism scientist
Laurent Mottron (2011, pp. 34-5) noted a bias in autism research towards undue negativity:
“researchers who study autism can display a negative bias against people with
the condition. For instance, researchers performing functional magnetic
resonance imaging (fMRI) scans systematically report changes in the activation
of some brain regions as deficits in the autistic group — rather than evidence
simply of their alternative, yet sometimes successful, brain organization.
Similarly, variations in cortical volume have been ascribed to a deficit when they
appear in autism, regardless of whether the cortex is thicker or thinner than
expected. When autistics outperform others in certain tasks, their strengths are
frequently viewed as compensatory of other deficits, even when no such deficit
has been demonstrated empirically.”
Autistic sociologist Damian Milton (2014) has written extensively on how autistic individuals
have been “frozen out” of the processes of knowledge production on autism. This in turn
undermines the “ethical and epistemological integrity” of autism research. It has meant the
focus has been driven by neurotypical norms and concerns; the assumptions of the scientific
paradigm may already exclude the possibility of seeing autism in a less negative light. Given
such worries, Walker (2012) endorses the trend to turn to a “neurodiversity paradigm”
instead of the “pathology paradigm” when conceptualising autism. This seeks to frame
autism in its wider social-ecological context, rather than by reducing it to biomedical deficit
compared to an elusive neurotypical norm.
This is important when it comes to conceptualising the dysfunctional element, and in turn the
target for intervention. On the medical model, the dysfunction is primarily conceived of as a
biologically-based dysfunction internal to the autistic individual. By contrast, the
neurodiversity research paradigm frames the dysfunction in at least two different ways. The
first is as a two-way relational dysfunction between autistics and non-autistics. Whilst autism
has traditionally been associated with impairments in empathising, neurodiversity proponents
have reframed autistic social and communication problems in light of how the empathy deficit
goes both ways between the autistic and neurotypical populations (Chapman, 2019b; Milton,
2012). Moreover, autistic individuals have often noted that they can empathise with other
autistic individuals (Chapman, 2019b) and neurotypicals (Brewer & Murphy, 2016). Second, the
dysfunction may be a relational one between the autistic and their society, whereby society
has been ordered and designed in a way that devalues and disables the autistic form of life.
While this view does not deny that some limitations will primarily stem from the autistic
individual, neurodiversity proponents thus reconceptualise the dysfunction by looking at the
wider context. This has important implications for attempts to remedy social and
communication issues, shifting the emphasis and responsibility away from a sole focus on
Opposition to Therapy : ABA Case Study
Viewing autism as difference rather than dysfunction, and awareness of the social
model of disability, combine to raise important questions about what – if any –
therapy offered to an autistic child or adult is helpful and legitimate, as opposed to
harmful and oppressive. Within the neurodiversity movement there is a range of
perspectives on the value of therapy. The call for acceptance and acknowledgement
of strengths, and indeed for socio-cultural changes to make the world more autism-
friendly, need not negate the fact that many autistic people will seek or need
individually-targeted assistance and support. What is called into question, though, is
what kinds of support are truly helpful. Awareness of the neurodiversity movement’s
concerns require important questions to be asked by all who seek to intervene
therapeutically: What is the purpose of therapy? What is the nature of the therapy?
What are the outcomes – both intended and unintended?
In some cases, a broad consensus is easily achieved that a particular approach is harmful.
Medical-model and social-model sympathisers agree about the harm linked to certain
interventions, such as packing (Amaral et al., 2011), and holding therapy (Fleming et al.,
2015). Far more controversial are those therapies that some hold as beneficent and others
hold as harmful. Behaviourally based approaches to autism are probably where opinion is
most sharply divided. They have respectability in some quarters and are loathed in others,
and are the focus of particular attention from the self-advocacy movement. In the following
section we explore the controversy, propose areas of discussion that remain outstanding
and where some resolution could be found, before finally outlining the areas that remain
unchartered across all therapeutic domains
Although there are many delivery models within the field of ABA, the dominant form that is
often cited by the neurodiversity movement is the UCLA Model, based on the early work of
Lovaas and colleagues (Lovaas, 1987). Thus, Wilkenfield and McCarthy (2020) describe a
typical ABA programme as involving 25-40 hours per week of intense one-on-one therapy,
“in which the autistic child is rewarded for engaging in activities that make him more
normal” (p. 37). Lovaas-style Early Intensive Behavioural Intervention (EIBI) with regard to
autism is given particular examination both because it was the first to be widely
disseminated in North America, and because it is equated in the eyes of some with an
attempt to eliminate all signs of autism – even to the point of “recovery”. Lovaas’ reference
to his autistic patients as “not persons in the psychological sense” (1974, p. 76) reflects a
dehumanising view of autistic people.
Unlike some other approaches, the Lovaas model of
intervention is criticised for failing to accommodate autistic characteristics, instead choosing
to wage “all out war against them ” (Cohen, 1998, p. 106) . For those in the neurodiversity
movement who hold that autistic behaviours such as stimming are fundamental, that autism
is key to someone’s identity, and further that it confers a collective identity, this reads as
“all out war on the autistic person” or “all out war on autistic people”. We now tackle these
issues in turn, testing the validity of the allegations, and then propose areas in which there
may be room for mutual accommodation.
Evidence and Efficacy
A standard claim made by the behaviour analytic community is that theirs is a discipline
rooted in science and that it uses systematic evidence at its base - in contrast to other
interventions that are held be less evidence-based, and less rooted in scientific principles
(Shreibman, 2005, p. 133). This claim requires unpacking, because references to the
evidence base of ABA can mean two different things. One relates to how individual practice
is rooted in a cycle of continually-reviewed empirical testing. The other relates to what
large-scale intervention trials may or may not tell us about the comparative impact of ABA
versus other approaches. Both will now be addressed in turn.
ABA adopts an individual-based ‘before and after’ approach, in which stimuli and
consequences are systematically recorded and manipulated so that it is possible to identify
precisely what component of an intervention is helpful or impactful. It is argued that this
makes it possible to identify quickly if an approach needs to be adjusted. It puts the onus on
the teacher or therapist to make learning as motivating as possible for the learner, and to be
This was in an interview with Psychology Today in 1974. The full quote is as follows: “You see, you start pretty
much from scratch when you work with an autistic child. You have a person in the physical sense — they have
hair, a nose and a mouth — but they are not people in the psychological sense. One way to look at the job of
helping autistic kids is to see it as a matter of constructing a person. You have the raw materials, but l you have
to build the person.”
accountable for failures to learn. The additional advantage of such an approach, it is argued,
is that it can be individually-tailored, according to the particular circumstances of each
person – building on their passions and interests, and requiring the teacher/therapist to
make adjustments if the individual is struggling. The additional justification for the origins
of ABA is that it avoids the excesses of psychological interventions that are based on
subjective, unreliable/partial narrative-based accounts, overlaid with non-proven
psychoanalytic theories. The rejection of mentalism in favour of reliance on the observable
and measurable has an integrity and rigour at its roots.
The ideal of individually-tailored practice is, however, undermined significantly in the case
of those EIBI interventions that adopt rigid curricula and are dominated by an aim to
remove or mask someone’s autistic behaviour. And more generally, there is a danger within
all BA practice that the emphasis given to precision data and research validity takes
precedence over any incidental change that has not been directly observed or measured.
The need for data integrity may be a safeguard against charlatans or wasteful and
ineffective therapeutic interventions, but it may also overlook influences and outcomes that
are not considered by the therapist as relevant and/or measurable at the point of
intervention. By avoiding subjective measures or qualitative research methods, limits are
placed on what areas of enquiry and knowledge can be explored within behaviour analytic
discipline. This inhibits curiosity in areas beyond the observable and measurable, which in
turn may limit the practitioner’s orientation towards any hidden or longer-term influences
of the therapy – some at least of which may be adverse.
The second pillar of the BA community’s claim to being evidence- based rests on
intervention trials that purport to indicate effectiveness, often in comparison with
alternative interventions (Foxx, 2008). But the status of this evidence is widely disputed. Just
as the original Lovaas studies were subjected to methodological criticism, so have
subsequent trials been noted for falling short of optimum design (Reichow et al., 2018). Thus,
Dawson’s influential attack “The Misbehaviour of Behaviourists” was strongly linked to a
rights-based position that standards of research on autistic subjects should be as robust
they are for anyone else (Dawson, 2004). She also pointed out, as others have done, that
relatively weak evidence has been promoted as if it were conclusive, in order to boost a
commercially-driven business approach to autism therapy, or - one could add - a
competitive approach from research centres driven by the quest for academic reputation.
For some critics, all forms of ABA, not just the dominant EIBI approach, are likely to be
rejected by virtue of the intellectual basis of behaviourism (Sparrow 2018). The highly-
publicised excesses of some of its applications further taint its reputation – for example
attempts to reduce homosexual behavior. Some autistic self-advocates hold that EIBI is
equivalent to this. In attempting to steer someone away from their core identity because it
clashes with restricted and oppressive social norms, the issue is that ABA seeks to normalize
autistics instead of helping them achieve self-actualization.
ABA is inherently vulnerable to the charge that it is a vehicle to enforce social conformity.
The profession defines itself as operating within social norms: “using scientific principles and
procedures discovered through basic and applied research to improve socially significant
behaviour to a meaningful degree” (Association of Professional Behaviour Analysts, n.d.).
Key texts refer to how ABA can steer behaviour to what is “socially important”, aiming to
influence behaviors for the “better state of society” (Baer, Wolf & Risley, 1968, p. 91). This
then implies an underlying belief that an individual’s own interests can be furthered by
adhering to what wider society holds to be important, and/or that it is permissible to deny a
person their own preferences and interests in order to fit in with the wider interests of
society as a whole. While some influential behaviour analysts have given consideration to
the issues around social validity, (Baer & Wolf, 1987; Wolf, 1978) few subsequent
approaches to ABA with regard to autism have fully addressed this challenge as matter of
A number of pressing questions arise here. Who decides what is socially important? And in
any case, should a helping professional aim to promote what is socially significant, or rather,
aim to address particular issues that are of distress and relevance to an individual? If people
are socially anomalous, what is the rationale for forcing social compliance? Clearly, what is
judged to be socially significant or socially important is open to wide interpretation, and is
very far from being self-evident. In the eyes of many autistic critics, ABA has entailed a
slavish and unquestioning adherence to norms dictated by a neurotypical (NT) majority.
One response from within the discipline of ABA has been to imply that the needs of the
individual are best met if they conform to behaviours that NT society holds as significant.
While this may be true in some cases (addiction or criminality, for example,) it is far more
questionable regarding autism. Shreibmann (2005, p. 238) justifies an attempt to redirect
stims as a way of promoting wellbeing via social acceptance - reducing the risk of being
stigmatised and ostracised. Yet in the case of some aspects of typical autistic behaviour,
such as stimming, the counter argument is that these are neither disadvantageous to society
nor to the individual – they are simply different or actually vital to the autistic person’s
identity and wellbeing (Rudy,2018). Autistics are forced to comply in a way that is not only
actively damaging to them, but also does not in any case bring about societal improvement.
Indeed, from the autistic perspective, this makes society worse, by reinforcing the
stigmatization of autistic behaviours more broadly. In response to the suggestion that a
reduction in atypical behaviour may minimise the stigmatisation or bullying of the
individual, the counter-argument holds that bullying is the aberrant behaviour and that the
person in need of a behavioural programme is the bully. As such, ‘all out war on autism’
runs counter to the idea that behaviour change – if it is required – should not be uni-
directional, but, rather, entail reciprocal adaptations to be made by the NT majority.
Trauma and Wellbeing
In making the case for the use of ABA, comparisons have occasionally been drawn between
treating autism and conducting surgical operations to address a serious illness, for example,
Frederikson (2013). The language used points to a fundamental category-error identified by
self-advocates who argue that their identity and way of being is in no way analogous to a
serious operable condition. The error gives room for therapists to justify uncomfortable
features of their practice: If autism is deemed to be pathological, then ‘treatment’ that is
considered necessary may be painful and distressing yet justifiable by a longer-term goal of
‘recovering’ from the pathology. But if a condition is not considered equivalently harmful or
dangerous, then any uncomfortable or distressing features of the intervention are
questionable at best. And, further, if the condition is experienced as integral to a person’s
identity, then intervention to eliminate it has significant moral implications. Thus Dawson
(2004) states that “failing to detect that people with differences have abilities and worth,
because they fail to be like those self-designated as exemplifying capability and worth, is a
denial of basic human rights”. The neurodiversity movement has also aligned itself with
human rights more broadly, for example the slogan “Autistic rights are human rights”
The internet is awash with forums and first-person accounts of autistic people who describe
their experience of receiving ABA as traumatic and harmful (Asasumasu, 2013; Kirsty, 2017).
The trauma relates not only to how ABA was practised on them – for example the use of
aversives that characterised early ABA programs - but also to the feelings of being powerless
and regarded as essentially flawed by virtue of being autistic. At the same time, it has been
found that adverse effects have been seriously understudied in research on autism
interventions, so there is little evidence to the contrary (Bottema-Beutel et al 2020).
Preliminary – and highly controversial – research by Kupferstein (2018) indicated that the
incidence of PTSD was far greater amongst those who had received ABA as children than
amongst those who did not. The Kupferstein study was was quickly criticised by Leaf et al
(2018) for its methodological limitations. In turn Chown et al (2019) pointed out that the
vested interests of Leaf et al and endorsed the Kupferstein findings as requiring further
investigation. In line with Chown et al’s suggestion, a recent thematic analysis of the
perspectives of twelve autistic adults who had experienced ABA for more than six months
found that they tended to associate ABA both with trauma and with a long term negative
impacts on their sense of self (McGill and Robinson 2020). While this is based on a small
sample, it is worth noting that more robust research found that the extent to which people
camouflage signs of autism in order to emulate being NT correlates with suicidality (Cassidy
et al., 2019) and depression (Cage et al., 2017) among autistic adults. While correlation does
not equate with causality, it is also notable here that the contrasting practice of ‘autism
acceptance’ - which refers to accepting autism as an intrinsic and valuable part of the
individual, and therefore adjusting environments and expectation accordingly - is not
associated with trauma. In fact, as an alternative to trying to treat autism, autism
acceptance has recently been found to help increase wellbeing in both autistic individuals
(Cage et al., 2017) and family members (Da Paz et al., 2018).
Autonomy and Identity
Autonomy, conceived as the capacity for self-governance, is widely seen as a fundamental
part of the good human life. Another reason ABA is taken to be permissible is that it
enhances autonomy, which is often taken to be diminished in the autistic population due to
the effects of cognitive processing differences associated with the autistic population.
Furman and Tuminello (2015), for instance, argue that ABA is vital for allowing autistic
individuals to be able to live good lives in large part because it can increase autonomy. In
their words, ABA may:
“assist the child with autism to develop his scientific reasoning skills—by breaking problems
down into manageable units—until he can autonomously discover/ justify truths inductively
or deductively.” (p. 260)
In line with this, some parents of multiply-disabled, non-verbal autistic individuals have
argued that ABA has led to significantly increased efficacy in individual autonomy. Examples
include the teaching of basic skills that mean less external support is required, by redirecting
and reducing behaviours that are perceived as dangerous, such as extreme self-injury or
pica, and by promoting communication that in turn increases the range of opportunities to
express preferences and make choices (Abaa4all, n.d.). The following observation by Wendy
Webster (2020) about Simon, her autistic son’s learning, in contrast to that of his NT siblings
Oliver and Lucy, is typical of how many parents feel:
The emphasis was always on teaching Simon skills that would help him have more
control in his life—either by being better able to communicate his wants and needs
or by being more independent.
Thanks to ABA, I have learned to be a much better teacher so that I could teach
Simon some of the same skills that I had taught Oliver and Lucy, such as buttoning
his shirt and tying his own shoelaces. In our world, ABA has opened lots of doors
and I do think that Simon has an infinitely better quality of life thanks to it. I don’t
feel that we’ve tried to make him more neuro-typical via ABA. On the contrary, the
more we’ve taught him, the more he’s been able to develop his own areas of
interest. For example, he can search for highly unusual (to me) videos on Google.
In contrast others have argued that ABA may in fact unduly subvert autonomy, when
directing a child’s behaviour away from their passions or their stims. Perhaps the strongest
reiteration of this criticism comes from a recent article by Wilkenfeld and McCarthy (2020).
They argue that what is at stake is a ”systematic violation of fundamental tenets of bioethics
– those of justice and nonmaleficience, and of autonomy”. While they recognise that
children do not have capacity at a young age to be authoritative decision-makers on several
issues of importance, they suggest that ABA offers a particular violation because it seeks to
alter the formation of a person’s identity in a way that is misaligned with their underlying
preferences, passions and pursuits: “ABA coercively closes off certain paths of identity
formation. This undermines children’s autonomy, and it fails to provide the kind of
autonomy support necessary for good caregiving” (Wilkenfield & McCarthy, 2020, p.47).
According to this view, then, what seems like an increase in autonomy might actually be a
decrease of autonomy in a more fundamental sense. Moreover, ABA is thereby at its most
harmful when it is at its most effective.
While this may be so in some cases, it is important to note that their critique is based on a
Lovaas-style model of EIBI, which in many cases has been superseded. Also, some have
questioned conflations between identity and autonomy in bioethical debates (McKenzie,
2020). In some cases, it may be that external influence on identity formation is not
necessarily autonomy-undermining. In the case where a child learns new skills or to avoid
self-injurious behaviours only, and is simultaneously supported in autism acceptance, it is
possible that autonomy is advanced without subverting individual identity. Still, even if the
case forwarded by Wilkenfield and McCarthy is overstated, it is vital to recognise that
teaching skills, if done in a way that relies on total compliance, may itself undermine the
development of autonomy in other ways. For instance, some autistic people who have
experienced ABA have reported that they were implicitly “taught to be compliant without
question” and that this “ultimately made it more difficult for them to say “no” as adults,
even when that “no” would protect them in a potentially dangerous situation” (Gardner,
2017). In line with this, Sandoval-Norton (2019) recently reviewed literature concluded that
ABA often “created prompt dependent adults who lack in self-motivation and self-esteem”
(p. 4). While ABA may be help develop autonomy-enhancing skills for some individuals, it
has often been applied in a way that may undermine autonomy.
Can ABA be Reformed?
It is important to recognise that the context in which ABA is practised may influence how it
is delivered and to what purpose. While some ABA practitioners do aim at extinguishing
autistic behaviours, others focus on issues that they believe can improve quality of life –
such as tooth-brushing, self-management of medication, and learning to use augmentative
communication devices. These are in keeping with what Wilkenfeld and McCarthy (2020)
refer to as developmental skills which they acknowledge are not inherently problematic as
therapeutic targets (p.56.), and which others (including ABA-sceptics) have maintained are
actively helpful (Pantazakos, 2019). It is also probably fair to say that while some parents
and practitioners might seek the shorthand outcome of ‘recovery’, others are aiming to help
autistic and learning disabled children acquire some core learning strategies - strategies
through which they can achieve in areas that will be helpful to them, without undermining
their self-esteem or attempting to change their diagnostic category.
Given that the practice of ABA is highly variable in terms of goals, context, approaches and
client group, some members of the neurodiversity movement have recognised that there
are relevant distinctions in how it is applied across the autistic population. Murray (2020)
refers to “British version nicey-nice ABA”. This falls short of an endorsement, but it does
recognise as relevant the distinction between different ways of implementing ABA. Despite
the critics’ focus on attempts to extinguish stimming or special interests, these examples are
not necessary features of all ABA approaches to autism, where passions and interests are
encouraged and a focus on communication – spoken or through alternative means such as
signing and PECS - works with, rather than against, the child.
Just as ABA practice is not homogenous, there is also a dramatic heterogeneity within
autism. The possibility that Autism Spectrum Disorders is an umbrella of varying subgroups
(Waterhouse & Gillberg, 2014); Muller & Amaral, 2017) reinforces the adage that in any kind
of therapeutic or educational approach “one size does not fit all”. Statistical trends in large
group studies may hide valuable data about diverse individual responses across all types of
therapy. To assert that ABA is harmful for some autistic people does not exclude the
possibility that it is helpful for some other autistic people, perhaps especially some of those
whose autism intersects with additional disabilities (although equally it could be more
harmful for some specific intersections). Any treatment can be extremely harmful when
forced or applied to the wrong trait, individual, or population, but beneficial for a small
number of people when applied appropriately, and there is not yet sufficient evidence to
conclude that ABA is necessarily harmful as such.
However, some autistic self-advocates have criticised attempts to distinguish between
“good” and “bad” kinds of ABA, arguing that all applications express the same underlying
attitudes to normalisation (Asasumasu, 2017) or because of issues in the conceptual
foundations of behaviourist principles (Sparrow, 2018). It is also notable that, at the time of
writing, research just published in the Journal of Applied Behaviour Analysis still classifies
autistic stims including as hand flapping as “problem behaviours” (Frank-Crawford et al.,
2020). The multiple testimonies of those for whom the experience of ABA has been
extremely negative therefore need to be taken very seriously by the ABA community. The
evidence of damage cited previously (Cassidy et al., 2019; Kupferstein, 2018)) – not to
mention the benefits associated with autism acceptance – all point to the need to engage
with self-advocates about the features of receiving ABA that they find particularly
distressing. While the Kupferstein sample is non-random, this does not undermine the
existence of the distress also attested to by multiple autistics who have experienced ABA
Importantly, Behaviour Analysts can look to discussions within their own discipline to
recognise ways in which subjective views and priorities require attention. Wolf wrote as far
back as 1978 that “… we must develop systems that allow our consumers to provide us
feedback about how our applications relate to their values…” (Wolf, 1978, p. 213). This is an
important statement made by an influential behaviour analyst, before the work of ABA
became so linked with a particular type of autism intervention. Likewise, the ABA literature
has offered reflections around the ethical issues of autonomy, consent and best interests
when the clients are children and/or people with limited capacity (Goldiamond, 1974). ABA
practitioners and academics should also be willing to take on board the insights of the social
model of disability, and be willing to critique some of the assumptions underlying a drive to
reduce autistic symptoms. This too lies within the tradition of reflective ABA practice: “A
proper appreciation of context always implies that we are not merely studying or managing
it, but also are part of it and therefore are being managed by it” (Baer et al.,1987)
Drawing these arguments together, then, a summary of where conversation across the
divisions may start is as follows. If some ABA clinics and practices adopt an approach that is
hostile to autism and to autistic people, by attempting to camouflage their autistic
difference, then the neurodivergent (ND) advocates are unequivocally right to protest. For
ABA proponents to establish a case for moral permissibility, ABA practice needs to focus on
therapeutic targets that are considered relevant to and -- where capacity allows -- by the
individual. This reflects the broader point made elsewhere (e.g. Bovell, 2020) that it is not
autism in toto that should be targeted, but rather specific issues that autistic people
themselves would like help with. Where parents and family members are involved in
decisions affecting children and adults with intellectual disabilities, the same principles
apply. Bearing all this in mind, if ABA proponents can show that ABA can increase well-being
and autonomy without causing trauma for some subsection of the autistic community, then
it will have shown that ethical reform is possible and that ABA is not just morally permissible
but also a significant good in some cases. But if such evidence is not produced, then working
towards abolition in favour of alternatives seems to be the only morally permissible option.
Implications for all Autism Therapies
It should be stressed that several of the neurodiversity movement’s concerns about ABA
actually apply to all therapies that target the individual, insofar as they chiefly operate
within the medical model of disability. Most of the helping professions focus on individuals,
regardless of any broader social critique that their members may have.
“If any disvalue of ASD stems from societal forces, placing the full burden of
rectifying this situation on autists in unjust” (Wilkenfield & McCarthy, 2020, p. 53)
This comment concerning ABA is therefore of broader significance. It does not invalidate the
potential value of therapy, but it raises important questions about what kinds of therapy are
helpful, for whom, and for what purpose, and it takes forward the principle that NT and
neurodivergent people should be engaged in a project of mutual and reciprocal
conversation and behaviour change. All participants in the conversation need to recognise
that different goals may apply to different individuals, in different contexts. Within this, it
will be important for the ND movement to engage more closely with parents about what
they may wish for their children. While “recovery” may be held out as a possibility to some
parents, many others seek therapeutic intervention as something that can promote their
child’s welfare without taking away their autism or indeed their intellectual disability.
At the same time, comparative intervention trials should move away from testing global
targets, and instead explore specific goals that are both less controversial and less general;
for example, the acquisition of expressive communication tools (spoken or otherwise), or
learning to tolerate nutritious foods. Such interventions are unlikely to be necessary across
an entire cohort of autistic children or adults, but will be relevant for particular subgroups
who have distinct challenges in these kinds of areas.
Practitioners and researchers, in collaboration with autistic people and/or their advocates,
therefore need to identify and adopt relevant measures when conducting comparative
research trials. They need to move away from standardised measures such as autism
symptoms that dominate intervention studies (Provenzani et al., 2019), and tackle the
under-representation of autistic people with intersecting learning disabilities in the bulk of
autism research (Russell, et al., 2019). Likewise, longitudinal markers regarding economic
status or type of living arrangement (Howlin et al., 2013) need to be questioned in terms of
their relevance, in contrast to other outcomes that may be more accurate as a measure of
quality of life or wellbeing across the autism spectrum.
This in turn opens up a wider question about what kinds of outcomes are truly positive as
opposed to merely fitting in with social conventions. The whole terrain of what kinds of lives
autistic people prefer, want and need is under-researched and under-discussed. Given the
heterogeneity of what is practical and desirable, according to the kinds of attributes and
additional issues facing a person (including the DSM “specifiers” about severity,
communication and intellectual ability) there will need to be a whole matrix of potential
signifiers for a good autistic life, defined in terms decided by the autistic individuals - either
expressly -- or by advocates who know them well.
While such shifts would be welcome, at this point in time, whether or not ABA can be
rehabilitated in the eyes of its critics remains to be seen. Currently, one of the co-authors
veers towards abolitionism, while the other veers towards reform. Regarding therapy more
broadly, only further enquiry can establish the impact of contrasting approaches in terms of
harms and benefits, where – crucially - these are framed as far as possible in terms of
perspectives on wellbeing described and observed among autistic people themselves and
those close to them, and taking into account the potential for significant variation across a
range of different intersecting disabilities.
Abaa4all. (n.d.) Family Stories. Retrieved November 20, 2020, from
Amaral, D. Rogers, S, J. Baron-Cohen, S. Bourgeron, T. Caffo, E. Fombonne, E. …. & van
der Gaag, R. J. (2011). Against le packing: A consensus statement. Journal of the
American Academy of Child and Adolescent Psychiatry, 2011 Feb: 50 (2), 191-2 doi:
American Psychiatric Association. 2013. Diagnostic and Statistical Manual of Mental
Disorders. (5th edn.) Washington, DC: APA.
ARGH (2014). T-shirts worn at Autism Rights Group Highland (ARGH) Autism and Ethics
Conference held on 16 May, Inverness. Further information available from:
Armstrong, T. (2015). The myth of the normal brain: Embracing neurodiversity. AMA Journal
of Ethics, 17(4), 348-352. doi: 10.1001/journalofethics.2015.17.4.msoc1-1504.
Asasumasu, K. (2013) The cost of indistinguishability is unreasonable. Radical
Neurodivergence Speaking. http://timetolisten.blogspot.com/2013/09/the-cost-of-
Baer, D. M., Wolf, M. M., and Risley, T. R. (1987) Some still-current dimensions of applied
behaviour analysis. Journal of Applied Behavior Analysis. 20 (4), 313-327
Barnbaum, D. R. (2008). The ethics of autism: among them, but not of them. Bloomington:
Indiana University Press.
Barnes, E. (2016). The Minority Body: A Theory of Disability. Oxford: Oxford University Press
Bannerman, D. J., Sheldon, J. B., Sherman J. A., and Harchik, A. E. (1990). Balancing the Right
to Habilitation with the Right to Personal Liberties: The Rights of People with
Developmental Disabilities to Eat Too Many Doughnuts and Take a Nap. Journal of
Applied Behavior Analysis 23, (1) 79-90
Booth, J. 2016. Autism Equality in the Workplace: Removing Barriers and Challenging
Discrimination. London: Jessica Kingsley Publishers.
Bottema-Beutel, K., Crowley, S., Sandbank, M., & Woynaroski, T. G. (2020). Adverse event
reporting in intervention research for young autistic children. Autism, 1362361320965331.
Botha, M., & Frost, D. M. (2020). Extending the minority stress model to understand
mental health problems experienced by the Autistic population. Society and
Mental Health, 10(1), 20–34. https://doi.org/10.1177/2156869318804297
Bovell, V. (2020) Is there an ethical case for the prevention and/or cure of autism? In:
H., Rosquvist, N. Chown, N. & A. Stenning, (eds) Neurodiversity Studies: A New Critical
Paradigm. (pp. 39-540. New York: Routledge.
Brewer, R. & Murphy, J. (2016). People with Autism Can Read Emotions, Feel Empathy.
Spectrum News 12 July 2016. Retrieved: June 29th, 2020 from
Cassidy, S. A., Gould, K., Townsend, E., Pelton, M., Robertson, A. E., & Rodgers, J. (2020). Is
Camouflaging Autistic Traits Associated with Suicidal Thoughts and Behaviours? Expanding
the Interpersonal Psychological Theory of Suicide in an Undergraduate Student Sample.
Journal of Autism and Developmental Disorders, 50(10), 3638–3648.
Cage, E., Di Monaco, J., & Newell, V. (2018). Understanding, attitudes and dehumanisation
towards autistic people. Autism 23(6) pp. 1373–1383
Cohen, S. (1998). Targeting Autism: What We Know, Don’t Know and Dan Do To Help Young
Children with Autism and Related Disorders. Berkeley: University of California Press.
Chapman, R. (2019a). Neurodiversity and its discontents: Autism, achizophrenia, and the
social model in S. Tekin, & R. Bluhm (Ed.) The Bloomsbury Companion to the
Philosophy of Psychiatry, (pp. 371-389). London: Bloomsbury. Chapman, R. (2019b).
Autism as a form of life: Wittgenstein and the psychological coherence of Autism.
Metaphilosophy 50(4): 421-440. https://doi.org/10.1111/meta.12366
Chapman, R. (2019b) The reality of Autism: On the metaphysics of disorder and
diversity. Philosophical Psychology. 33(6), 799-819.
Chapman, R. (2020). Neurodiversity, disability, wellbeing”. In: H. Rosquvist, N. Chown, & A.
Stenning, A. (Eds.), Neurodiversity Studies: A New Critical Paradigm. (pp. 57-72). New York:
Chown, N., Hughes, E., Leatherland, J., & Davison, S. (2019). Response to Leaf et al.’s critique
of Kupferstein’s finding of a possible link between applied behaviour analysis and post-
traumatic stress disorder. Advances in Autism, 5(4), 318–318. https://doi.org/10.1108/AIA-
Dawson, M. (2004). The misbehaviour of behaviourists: Ethical challenges to the autism-
ABA industry. Retrieved 29 June, 2020, from:
Da Paz, N. S., Siegel, B., Coccia, M. A., & Epel, E. S. (2018). Acceptance or Despair? Maternal
Adjustment to Having a Child Diagnosed with Autism. Journal of Autism and Developmental
Disorders, 48(6), 1971–1981. https://doi.org/10.1007/s10803-017-3450-4
Devita-Raeburn, E. (2016). The controversy over Autism’s most common therapy.
Spectrum News. Retrieved 29 June, 2020 from
Dyck E., & Russell G. (2020) Challenging psychiatric classification: Healthy Autistic
diversity and the neurodiversity movement. In: S. Taylor S., & A. Brumby
(Eds.) Healthy Minds in the Twentieth Century. Mental Health in Historical
Perspective (pp. 167-187). Palgrave Macmillan, Cham
Feinstein, A. (2010) A History of Autism: Conversations with the Pioneers. Oxford: Wiley-
Fleming, B., Hurley, E., and the Goth (2105) Choosing Autism Interventions: A research-
Based Guide. Hove: Pavilion Publish and Media Ltd.
Foxx, R. (2008) Applied behavior analysis treatment of autism: The state of the art. Child
and adolescent psychiatric clinics of North America, 17, 821-834
Frederiksen, G. (2013) quoted in Robinson, F. (Director) Autism: Challenging behaviour
[Television documentary] BBC 4, November 2013. Retrieved from Youtube:
Furman, T. M., & Jr, A. T. (2015). The Deep Impact of Applied Behavior Analysis for Children
with Autism Spectrum Disorder. Philosophy, Psychiatry, & Psychology, 22(4), 271–274.
Gardner, F. (2017) First-Hand Perspectives on Behavioral Interventions for Autistic People
and People with other Developmental Disabilities. Autistic Self-Advocacy Network.
Goldiamond, I. (1974). Toward a constructional approach to social problems, Behaviorism 2,
Howlin, P., Moss, P., Savage, S. and Rutter, M. (2013). Social outcomes in mid- to later
adulthood among individuals diagnosed with Autism and average nonverbal IQ as
Children. Journal of the American Academy of Child and Adolescent Psychiatry, 52 (6),
572–581. https://doi.org/10.1016/j.jaac.2013.02.017 DOI:
Kirsty (2017) ABA’s Harmful Message to Children: You can’t say “no”. Autistic Christian
Leaf, J. B., Ross, R. K., Cihon, J. H., & Weiss, M. J. (2018). Evaluating Kupferstein’s claims of
the relationship of behavioral intervention to PTSS for individuals with autism. Advances in
Autism, 4(3), 122–129. https://doi.org/10.1108/AIA-02-2018-0007
Lovaas, O. I. (1974)) interviewed by Paul Chance, Psychology Today 7 (January);
accessed 10 August 2019, accessible at:
(1987). Behavioral treatment and normal educational and intellectual
functioning in young autistic children. Journal of Consulting and Clinical Psychology,
55, 3-9. https://doi.org/10.1037/0022-006X.55.1.3
McGill, O., & Robinson, A. (2020). “Recalling hidden harms”: Autistic experiences of
childhood applied behavioural analysis (ABA). Advances in Autism, ahead-of-print(ahead-of-
print). https://doi.org/10.1108/AIA-04-2020-0025“About Behavior Analysis - Association of
Professional Behavior Analysts”. n.d. Retrieved 29 June 2020 from
Mottron, L. (2017). Should we change targets and methods of early intervention in autism,
in favour of a strengths-based education? European Child & Adolescent Psychiatry.
26, 815–825. https://doi.org/10.1007/s00787-017-0955-5
Milton, D. and Sims, T. 2016. How is a sense of well-being and belonging constructed in the
accounts of autistic adults? Disability and Society, 31 (4),520-534.
Milton, D. E. M. 2012. On the ontological status of autism: The double empathy problem.
Disability and Society 27, 883–887. https://doi.org/10.1080/09687599.2012.710008
Milton, D. E. 2014. Autistic expertise: A critical reflection on the production of knowledge in
autism studies. Autism, 18(7), 794–802.
https://doi.org/10.1177/1362361314525281Muller, R-A., & Amaral, D. G., (2017) Editorial:
Time to give up on Autism Spectrum Disorder? Autism Research, 10 (1), 10-14.
Murray, D. (2020) From Protest to Taskforce. In S. K. Kapp (ed.) Autistic Community and
the Neurodiversity Movement Stories from the Frontline, 277-285. Palgrave Macmillan.
Oliver, M. (1990) The Politics of Disablement. New York: Macmillan.
Provenzani U, Fusar-Poli L, Brondino N, et al. What are we targeting when we treat autism
spectrum disorder? A systematic review of 406 clinical trials. Autism 24(2), 274-284.
Reichow, B., Hume, K., Barton, E. E., , Boyd B.A. (2018). Early intensive behavioral intervention
(EIBI) for young children with autism spectrum disorders (ASD). Cochrane Database of
Systematic Reviews (5). Art. No.: CD009260. DOI: 10.1002/14651858.CD009260.pub3
Renty, J. O., & Roeyers, H. (2006). Quality of life in high-functioning adults with autism
spectrum disorder: The predictive value of disability and support characteristics. Autism:
The International Journal of Research and Practice, 10(5), 511–524.
Rodogno, R., Krause-Jensen, K., Aschcroft, R. E. (2016) ‘Autism and the good life’: a new
approach to the study of well-being. Journal of Medical Ethics. 42(6), 401-408.
Rudy, L. J. (2018 December 20). Overview of stimming in autism. Verywellhealth.com.
Retrieved from https://www.verywellhealth.com/what-is-stimming-in-autism-
Russell, G., Mandy, W., Elliott, D. et al. (2019) Selection bias on intellectual ability in autism
research: a cross-sectional review and meta-analysis. Molecular Autism 10, 9 (2019).
Schreibman, L. (2005). The Science and Fiction of Autism. London: Harvard University Press
Sinclair, J. 1993. Don’t Mourn for Us. Our Voice 1(3). [Online]. Available from:
http://www.autreat.com/dont_mourn.html [Accessed 12 October 2017).
Singer, J. (1999). Why can’t you be normal for once in your life? From a ‘problem with no
name’ to the emergence of a new category of difference. In M. Corker & S. French
(Eds.), Disability discourse, (pp. 59-67). Maidenhead, UK: Open University Press.
Sparrow, M. (2018) But what about the good ABA therapists?
Waterhouse, L., & Gillberg, C., (2014). Why autism must be taken apart. Journal of Autism
and Developmental Disorders, 44 (7), 1788–92. doi:10.1007/s10803-013-2030-5
Wakefield, J. 1992. Disorder as Harmful Dysfunction: A Conceptual Critique of D.S.M IIIR's
Definition of Mental Disorder. Psychological Review 99, 232-247. doi:
Walker, N. 2013. Throw Away the Master’s Tools: Liberating Ourselves from the Pathology
Paradigm. [blog post] Retrieved from http://neurocosmopolitanism.com/throw-
Walker, N. (2014). Neurodiversity: Some Basic Terms & Definitions. [blog post]. Retrieved
Walker, M., & Mackenzie, C. 2019. Neurotechnologies, relational autonomy, and
authenticity. International Journal of Feminist Approaches to Bioethics 13 (1): 98–
Waltz, M. (2008), Autism = Death: The social and medical impact of a catastrophic medical
model of autistic spectrum disorders', Journal of Popular Narrative Media, 1(1), 13-
Wilkenfield, D. A. & McCarthy, A. M. (2020). Ethical Concens with Applied Behavior Analysis
for Autism Spectrum “Disorder”. Kennedy Institute of Ethics Journal, 30 (1), 31-69
Wolf, M. M. (1978) Social validity: The case for subjective measurement or how Applied
Behavior Analysis is finding its heart. Journal of Applied Behavior Analysis, 11(2) 203-