Originating in the UK and United States during the 1960s and 1970s, the disability rights movement rejects the analogy between disability and disease (Shakespeare, 2013; Skrentny, 2009). Rather, the disability rights movement holds that disabilities are a type of diversity, analogous to race and gender (Barnes, 2016). Thus, although it is common to hold that disabilities are bad-differences—the kind of thing that make one’s life worse (Swain & French, 2000), the disability rights movement holds that disabilities are mere-differences, akin to race, gender, or sexual orientation (Barnes, 2016). It is important to note that the disability rights movement does not deny that people with disabilities often have harder lives; however, they explain this fact by pointing to both individual and institutional ableistFootnote 1 bias (Amundson & Tresky, 2007).

The neurodiversity movement is analogously committed to a mere-difference account of autism (and potentially of other developmental disabilities as well). Proponents of the neurodiversity perspective hold that there is nothing inherently wrong with being autisticFootnote 2 (Dwyer, 2022; Singer, 1999). Rather, similar to race and gender, being autistic constitutes one of the many ways in which people are diverse (Dwyer, 2022; Singer, 1999). The neurodiversity perspective is thus consistent with mere-difference models of disability and opposes the medical model, which holds that autism is pathological (Chapman & Bovell, 2022; Dwyer, 2022; Kras, 2009).

It is important to note that the quality of life of individuals with autism does not constitute the primary point of disagreement between the neurodiversity perspective and the medical model. Rather, in those cases where individuals with autism face unique challenges, the two perspectives disagree about the causes of these unique challenges. The medical model holds that when there are decreases in quality of life associated with having autism, they are primarily caused by autism (Kapp, 2019; Kras, 2009). By contrast, the neurodiversity perspective holds that when there are decreases in quality of life associated with being autistic, they are primarily caused by an interaction between the autistic person and their environment (Chapman, 2019). For example, sensory sensitives to light or sound can make it more challenging for autistic people to navigate physical environments designed for neurotypical people (Belek, 2019; MacLennan et al., 2022). Likewise, the stigmatization of nonneurotypical behavior (e.g., hand-flapping) can make it difficult for autistic individuals to navigate common social contexts (Rothman et al., 2022).

The literature on neurodiversity often associates applied behavior analysis (ABA) with the medical model (Chapman & Bovell, 2022; Kapp, 2019; Krcek, 2013). At some level, this is unsurprising; behavior analytic interventions remain a foremost medically advised response to autism, including explicit support for behavior analytic interventions by the U.S. Centers for Disease Control (CDC, 2019), The American Academy of Pediatrics (e.g., Hyman et al., 2020), and the National Institute of Mental Health (NIMH, 2011). A number of states also mandate that medical insurance providers cover the costs of ABA therapy for children with autism (National Conference of State Legislatures, 2022).

Given that proponents of the neurodiversity perspective reject the medical model and given that ABA is often associated with the medical model, it should come as no surprise that ABA is often a target of criticism from proponents of the neurodiversity perspective. Critiques of ABA from the neurodiversity movementFootnote 3 have been most prevalent in the form of internet blog posts and on social media platforms, with apparently increasing prevalence over the past 5 to 10 years (e.g., Bascom, 2011, 2014; Devita-Raeburn, 2016; Leaf et al., 2021; Lynch, 2019; Ram, 2020). In light of concerns raised by the neurodiversity movement, peer reviewed research articles have also increasingly begun to articulate worries regarding the safety and ethics of behavior analytic research and practice involving autistic persons (e.g., Bottema-Beutel & Crowley, 2021; Dawson & Fletcher-Watson, 2022; Lai et al., 2020; McGill & Robinson, 2020; Schuck, Dwyer, et al., 2022a; Wilkenfeld & McCarthy, 2020).

Proponents of the neurodiversity perspective have raised a wide range of concerns regarding the use of ABA interventions with autistic clients. Many of these target specific ways of practicing ABA (e.g., the use of aversives; Devita-Raeburn, 2016), the number of hours of ABA involved in early intensive behavioral intervention (EIBI; Wilkenfeld & McCarthy, 2020), or the social validity of pivotal response treatment (Schuck, Dwyer, et al., 2022a). Other concerns involve putative long-term harm from the application of ABA therapy to autistic clients (e.g., posttraumatic stress disorder; Kupferstein, 2018; but cf. Leaf et al., 2018) and even suicidality (Mitchell et al., 2021; Ne’eman, 2021). These concerns largely revolve around the ways in which ABA interventions for autistic clients are implemented. These criticisms are thus compatible with the aim, embraced by some neurodiversity advocates, of “reforming” ABA (Chapman & Bovell, 2022; Schuck, Tagavi, et al., 2022b).

By contrast, a distinct theme in the neurodiversity literature aims to raise concerns about the aims or goals of the application of ABA interventions when working with autistic clients. This latter challenge raises concerns, not just about how the practice of ABA is implemented, but at the appropriateness of even designing ABA interventions for autistic clients. As such, this latter challenge, which we will call the abolitionist neurodiversity critique, rejects the possibility of “reforming” ABA and rather supports “working towards abolition” of ABA (Chapman & Bovell, 2022, p. 1532).

In perhaps the highest profile “success” of the abolish-ABA movement, on July 21, 2021, the National Council on Independent Living (NCIL) passed a resolution condemning ABA as “a harmful and abusive practice that the National Council on Independent Living (NCIL) opposes in all its forms” (National Council on Independent Living, 2021). Though NCIL’s anti-ABA resolution stands out for having been passed, a number of advocacy organizations continue to advocate for similar “bans” on ABA therapy (e.g., Autistic Collaboration n.d.; LGBTQNation, 2021).

Despite the increasing prominence of the neurodiversity movement, the literature on neurodiversity in behavior analytic publications remains in its nascent stages. The Association for Behavior Analysis International links to 19 behavior analytic journals.Footnote 4 Of the articles found in these 19 journals (excluding book reviews), neurodiversity is only mentioned in 5 (Catrone et al., 2022; Leaf et al., 2022; Najdowski et al., 2021; Veneziano & Shea, 2022; Walsh, 2011).Footnote 5 This compares unfavorably to an ongoing, four-article exchange in JAMA Pediatrics kickstarted by Dawson et al. (2022) and very unfavorably to the 2,120 books, articles, whitepapers, and theses returned by a Google Scholar search for “neurodiversity” limited to the year 2020.

Although still minimal overall, dialogue regarding neurodiversity within the behavior analytic community has been increasing over the past few years. For example, dialogue related to neurodiversity at behavior analytic conferences has been gradually increasing (e.g., Graber & Graber, 2018; Lerman et al., 2022; Murray, 2021; Rodriguez et al., 2022; Sundberg et al., 2022; Weiss et al., 2021), including presentations by proponents of the neurodiversity movement (e.g., Murray, 2021; Rossi, 2021; Sundberg et al., 2022) and autistic individuals (e.g., Bodkin, 2021; Gravino, 2022; Hall, 2022; Holcomb, 2021), as well as panel discussions in defense of ABA (Weiss et al., 2021). Furthermore, in recent years a small number of peer-reviewed research articles have been published responding to neurodiversity critiques of ABA (e.g., Gorycki et al., 2020; Leaf et al., 2018; Leaf et al., 2021).

Simultaneously, behavior analytic research has increasingly referenced the importance of elevating diverse voices in behavior analysis and creating a more inclusive culture within the field (Bodkin, 2021; Brodhead, 2019; Conners et al., 2019; Conners & Capell, 2020; Fong et al., 2017; Holcomb, 2021; Murray, 2021; Rossi, 2021; Wright, 2019). Nonetheless, at least within peer-reviewed publications and behavior-analytic journals, the behavior-analytic community has overall been conspicuously quiet on the implications of the neurodiversity movement for building a more inclusive culture.Footnote 6

This article seeks to expand the discussion of neurodiversity within behavior analysis. In particular, this article aims to introduce and assess the abolitionist neurodiversity critique of the aims of the clinical practice of ABA for autistic clients, and to consider concomitant ethical obligations relevant to the practice of ABA.

This article proceeds in three major sections. In the first section we will introduce the reader to the abolitionist neurodiversity critique and illustrate the force of the critique via three hypothetical case studies. In the second section we will use an analogy to the challenges posed by linguistic diversity in classrooms to answer the abolitionist neurodiversity critique. Finally, we will consider the broader ethical implications for use of ABA-based therapies for individuals with autism.

Although embracing reformist critiques of ABA is compatible with being a behavior analyst, the same cannot be said about embracing abolitionist critiques of ABA. Someone who believes that ABA-based intervention is morally problematic in a way that is beyond reform cannot, in good conscious, remain a practicing behavior analyst. Thus, whereas behavior analysts can, and should, take reformist neurodiversity critiques seriously, the abolitionist neurodiversity critique presents a unique challenge for behavior analysts. For this reason, and given the limited scope of a single article, we will focus exclusively on a abolitionist critique of ABA. Though our focus on the abolitionist neurodiversity critique precludes engagement with the many reformist concerns raised by the neurodiversity movement, there remains a pressing need for future articles that grow the discussion of neurodiversity in the ABA literature.

The Abolitionist Neurodiversity Critique

Introducing the Critique

The abolitionist neurodiversity critique highlights (at least) two common concerns about the aims of ABA. Each is grounded in the idea that, like race, gender, or sexual orientation, autism is a mere-difference. First, proponents of the neurodiversity perspective are likely to criticize treating neurotypical behavior as the benchmark for success: “[ABA] does not help the autistic person in question, it only makes the autistic person more ‘acceptable’ to others” (Schuck et al., 2022a, p. 12; see also, Chapman & Carel, 2022). Second, proponents of the neurodiversity perspective are likely to criticize treating autism (or engaging in neurodivergent behavior) as something that is inherently bad or problematic (e.g., Chapman, 2019; Chapman & Carel, 2022). In this section we will introduce three hypothetical case studies and consider how these concerns apply to each. In the following section we will introduce an analogy to guide our analysis of the three case studies. Within the framework of the analogy, we will then return to the case studies and analyze them anew.

In order to avoid the potential distraction caused by unnecessary details, the case studies are bare bones. There will be ways of filling in each case study where the behavior analytic practice looks very poor. Likewise, there will be ways of filling in each study where, setting aside the abolitionist neurodiversity critique, the behavior analytic practice looks quite good. No matter how one fills in the unspecified details of the case, the specified details should be enough to illustrate one way in which the abolitionist neurodiversity critique can apply to behavior analytic practice.

  • SELF-INJURIOUS BEHAVIOR: Avery is 13 years old and has been diagnosed with autism spectrum disorder (ASD). Avery engages in repetitive self-injurious behavior involving hitting her hands/fists to her forehead, resulting in chronic bruising and occasional open wounds on Avery’s hands and head.

  • HAND-FLAPPING IN SCHOOL: Alex is 13 years old, is in a general education setting, and has been diagnosed with ASD. Alex engages in repetitive hand-flapping that does not cause injury. Alex’s hand-flapping is incompatible with engaging in various tasks at school. Alex’s parents and teachers are concerned that her hand-flapping is interfering with her academic progress.

  • HAND-FLAPPING IN PRIVATE: Ace is 13 years old and has been diagnosed with ASD. Ace engages in repetitive hand-flapping in the privacy of their home but not in other contexts. Though Ace’s hand-flapping does not generally interfere with the achievement of social or academic milestones, Ace’s parents are nonetheless troubled by it and have requested intervention to eliminate hand-flapping.

The Abolitionist Neurodiversity Critique and SELF-INJURIOUS BEHAVIOR

The abolitionist neurodiversity critique applies in different ways to each of these hypothetical case studies. In SELF-INJURIOUS BEHAVIOR, the abolitionist neurodiversity critique has limited application. There are two ways in which one can argue against the use of ABA in the case of SELF-INJURIOUS BEHAVIOR: (1) it is wrong to use ABA in the treatment of self-injurious behavior (SIB) for persons with autism because it is wrong to treat SIB at all, in so far as the SIB is understood to be part of that person’s autism; or (2) SIB should be targeted for treatment because it is harmful, but methods other than ABA should be utilized.

Regarding the first concern, there seems to be little to no literature supporting the claim that treatment of SIB is altogether wrong. Rather, the literature from the neurodiversity movement discusses the importance of well-being and limiting harm to autistic persons in medical and therapeutic care (Shkedy et al., 2019; Wilkenfeld & McCarthy, 2020), consistent with the treatment of SELF-INJURIOUS BEHAVIOR to limit harm/injury. Concerns within the neurodiversity movement regarding intervention/treatment for SIB seem to center around the appropriate types and methods of treatment (e.g., Shkedy et al., 2019), rather than whether or not SIB should be treated.

In a prominent example, Shkedy et al. (2019) suggest that treatment approaches more commonly associated with clinical psychology, such as dialectical behavior therapy (DBT), should guide treatment. It should be noted that cognitive behavior therapy in general and DBT in particular share strong theoretical lineage and applied principles with ABA (Hayes et al., 2011). As the creator of DBT, Dr. Linehan, herself described it, “In essence, [the development of] DBT was a trial-and-error clinical effort based on the application of behavioral principles” (Linehan & Wilks, 2015, p. 97).

Shkedy et al. (2019) also states that the treatment of SIB for an autistic person should be focused on building that person’s ability to functionally communicate, tolerate frustration, and regulate their emotions, whether that person is autistic or neurotypical. A close reading of Shkedy et al. (2019) finds some acknowledgement that research in ABA involves attention to these goals, such as when Shkedy et al. (2019) briefly reference functional communication training (FCT; Carr & Durand, 1985). In fact, FCT is understood to be the most widely and commonly used, function-based behavioral treatment for problematic behaviors in ABA, across various demographics and treatment modalities (Ghaemmaghami et al., 2021; Kurtz et al., 2011; Tiger et al., 2008; Wacker et al., 2011, 2013). Nonetheless, Shkedy’s claim does not seem to constitute an all-out denial of effective treatments for SIB within ABA, but rather a claim that “the application of this research is lacking” (Shkedy et al., 2019, p. 3).

Thus, concerns within the neurodiversity movement surrounding the treatment of SELF-INJURIOUS BEHAVIOR seem to relate more to the forms and methods of ABA practiced, rather than a call to abolish all forms of treatment for SIB that could be consistent with ABA. Therefore, neurodiversity critiques of ABA in the treatment of SELF-INJURIOUS BEHAVIOR seem to be best understood from the standpoint of a “reformist,” rather than an “abolitionist” critique. Although an important issue for the broader conversation surrounding the teachings of the neurodiversity movement for practitioners of ABA, a deeper dive into the specific methods within ABA for the treatment of SIB is beyond the scope of this manuscript, which focuses on the abolitionist neurodiversity critique from a disability-rights perspective (e.g., whether or not it is appropriate to target certain repetitive behaviors in autism for intervention).

In summary, the notion that SELF-INJURIOUS BEHAVIOR is appropriate to target for treatment seems to be compatible with a disability-rights perspective and with the abolitionist neurodiversity critique. In the context of the hypothetical case example, autism is a mere-difference but injuries are bad, and Avery’s repetitive behavior is causing injury in this case. There is thus a straightforward justification for behavioral intervention that neither treats neurotypical behavior as normative nor treats autism as pathological (Ne’eman, 2021), but involves the application of evidence-based treatments that minimize the risk of harm.

In Avery’s case, best practice mandates that a functional assessment be conducted to identify the environmental variables evoking and/or maintaining the self-injurious behavior and that matched function-based interventions be implemented (Behavior Analyst Certification Board (BACB), 2014, 2020). These ethical practice guidelines are based on extensive research over the past several decades demonstrating the effectiveness and social acceptability of the FBA/FCT model (e.g., Beavers et al., 2013; Carr, 1977; Carr & Durand, 1985; Ghaemmaghami et al., 2021; Ingram et al., 2005; Iwata et al., 19821994; Kurtz et al., 2020; Northup et al., 1991; Wacker et al., 2011, 2013). Nonetheless, function-based interventions for Avery should not be based on assumptions of normativity of neurotypical behavior. For example, Avery’s SIB may be evoked by work demands that are shaped by productivity benchmarks normed to Avery’s neurotypical peers. If this is the case, it may be more appropriate to adjust productivity benchmarks than to focus on modifying Avery’s behavior.Footnote 7

Function-based interventions can work to reshape both the individual’s behavior and the environmental context so as to not unduly disadvantage or burden neurodivergent individuals like Avery. Function-based, behavior analytic interventions for SELF-INJURIOUS BEHAVIOR will aim to identify and remedy the ways in which the social and physical environment constitute barriers to Avery’s wellbeing.

SELF-INJURIOUS BEHAVIOR is important because it highlights the limits of the abolitionist neurodiversity critique. Independent of one’s views on neurodiversity, some behaviors are unquestionably harmful. As a result, even when taken at face value, the abolitionist neurodiversity critique is not sufficient to make the case for the abolition of ABA.

The Abolitionist Neurodiversity Critique and HAND-FLAPPING IN SCHOOL

The abolitionist neurodiversity critique finds the other two case scenarios more problematic. Consider HAND-FLAPPING IN SCHOOL. In this vignette, hand-flapping is a behavior of concern because it threatens to interfere with Alex’s academic progress. Two features of HAND-FLAPPING IN SCHOOL give rise to concern from the perspective of the abolitionist neurodiversity critique.

First, focusing on hand-flapping obscures the role of educational design decisions in Alex’s academic experiences. If the repetitive behaviors characteristic of autism were seen as normal variants of student behavior within the public education context, coursework and classrooms would likely be designed in ways that made hand-flapping compatible with achieving desired learning outcomes. Focusing exclusively on Alex’s repetitive hand-flapping as the target for intervention threatens to obscure the fact that Alex’s hand-flapping only interferes with academic progress because educational environments are designed in ways that privilege neurotypicality.

Second, the degree to which hand-flapping interferes with Alex’s academic progress depends in large part on the existence of a normative rate of academic progress. The abolitionist neurodiversity critique raises the question: How do we decide the rate at which Alex is expected to learn? Academic (and social) benchmarks are set largely by reference to Alex’s neurotypical peers, yet proponents of the neurodiversity perspective will likely view this emphasis on neurotypical performance as a form of prejudice. The neurodiversity perspective holds that autism is a type of diversity analogous to race or gender. Thus, from the neurodiversity perspective, just as it would be morally inappropriate to hold up maleness or whiteness as a standard to which everyone should aspire, it is likewise morally inappropriate to hold neurotypicality as the standard to which everyone ought to aim (e.g., Chapman & Bovell, 2022; Gibson & Douglas, 2018; Kirkham, 2017).

Though ABA intervention for Alex’s hand-flapping may treat neurotypical behavior as normative, it need not treat hand-flapping as pathological. Much will depend on the reason that Alex’s hand-flapping has been targeted for intervention. In the case study, hand-flapping was identified as a behavior of concern because it was interfering with Alex’s academic progress. Intervening with Alex’s hand-flapping on these grounds likely treats neurotypical behavior as normative but does not require an acceptance that hand-flapping is pathological.

Though the reasons given for intervention in HAND-FLAPPING IN SCHOOL do not pathologize hand-flapping, there are pathologizing justifications for intervening in HAND-FLAPPING IN SCHOOL. Justifications for intervention that pathologize Alex’s hand-flapping are morally problematic. Pathologizing autism or neurodivergent behavior is morally analogous to treating a person’s race or sexual orientation as a sickness that should be cured. From the perspective of the neurodiversity paradigm, just as viewing gayness as a sickness in need of a cure is wrong because it constitutes a failure to view gay people with dignity and respect, it is likewise wrong to treat autism as an appropriate target for “treatment” (Autistic Collaboration, n.d.; BACB, 2020).

The Abolitionist Neurodiversity Critique and HAND-FLAPPING IN PRIVATE

HAND-FLAPPING IN PRIVATE is the most vulnerable to critique. In this vignette, Ace’s parents have requested a behavioral intervention to reduce Ace’s hand-flapping. Ace’s hand-flapping occurs in the privacy of their home and does not clearly contribute to any stigma Ace might face as an autistic individual. Likewise, Ace’s hand-flapping does not seem to interfere with the achievement of social or academic goals. It is thus unclear why the parents would request intervention for hand-flapping that occurs in the privacy of their home unless they either view Ace’s hand-flapping as wrong or believe that hand-flapping expresses an underlying and pathological trait.

Yet the idea that hand-flapping is problematic in-and-of-itself is an instance of treating neurodivergent behavior as pathological, i.e. as a behavior for which reductive clinical intervention is warranted. Proponents of the neurodiversity perspective will note that, just as it would be morally inappropriate to view a person’s gender as being in need of “recovery,” it is morally inappropriate to view “recovery” from autism as an appropriate goal.Footnote 8 Likewise, just as it would be morally inappropriate to conceptualize being gay in terms of deficits, it is morally inappropriate to conceptualize having autism in terms of deficits (e.g., Chapman & Bovell, 2022; Gibson & Douglas, 2018).

From a Behavior Analytic Perspective, is Autism a Mere-Difference?

The most straightforward response to the abolitionist neurodiversity critique is to reject its central premise, i.e., to reject the claim that autism is a mere-difference. If autism is a bad-difference—the type of thing that by its nature makes a person’s life worse, then there is nothing problematic with treating autism as a pathology.

This is not, however, a response the behavior analyst can give. Autism is characterized by patterns of behavior and person–environment (social) interactions. In particular, individuals with autism are identified by the psychiatric/medical community as having, “persistent deficits in social communication and social interaction across multiple contexts” (American Psychiatric Association [APA], 2013, p. 50) and “restricted, repetitive patterns of behavior, interests or activities” (APA, , 2013, p. 50). Both criteria center on neurodivergent behaviors.

From a behavior analytic perspective, behaviors are not explained via internal mechanisms. Rather, behavior is the result of organism–environment interactions (Alberto & Troutman, 2013; Cooper et al., 2020; Skinner, 1965, 1981, 2019). In virtue of this focus on observable behavior, “The science of behavior analysis has made discoveries that have proven useful in addressing socially important behavior such as drug taking, healthy eating, workplace safety, education, and the treatment of pervasive developmental disabilities (e.g., autism)” (Association for Behavior Analysis International [ABAI], 2022). On the behavior analytic view, having autism is thus not wrong or pathological. This is exactly the view endorsed by the neurodiversity perspective; as Leadbitter et al. (2021) describe it, autism “arises from the interaction of a non-standard individual and an unaccommodating environment” (p. 2). Given that autism is defined behaviorally, the behavior analyst’s commitment to viewing behavior as an interaction between an organism and an environment should lead them to embrace the neurodiversity perspective of autism as a mere-difference, rather than a bad-difference (Graber & Graber, 2018; Rodriguez, 2022).

Responding to the Abolitionist Neurodiversity Critique

Additive and Subtractive Approaches to Linguistic Diversity

In what follows, we will draw an analogy to the longstanding debate over the appropriate approach to linguistic diversity in the classroom. Although there is much that can be learned from this debate, the analogy is nonetheless limited in important ways. For example, linguistic diversity involves a limited topography of behavior. By contrast, neurodivergent behaviors span a much wider range of behavioral topographies. Nonetheless, the literature surrounding linguistic diversity in the classroom tackles the thorny question of the appropriate approach to shaping behavior when, (1) students (or clients) have a behavioral repertoire that allows them to access reinforcement in some environments but not others; and (2) the behavioral repertoire students (or clients) possess is (wrongly) stigmatized. As the abolitionist neurodiversity critique highlights, this is exactly the situation behavior analysts find themselves in with regard to their clients. Thus, although the analogy to the debate over linguistic diversity in the classroom has important limitations, lessons from that literature are immediately and broadly applicable to serving autistic clients in ABA.

An analogy to linguistic diversity in the classroom can help us think more clearly about the concerns raised in the preceding discussion. There is a sizable academic literature that tackles the question of how best to approach teaching linguistically diverse students (e.g., Cummins, 1997; Delpit, 2006; Flores & Rosa, 2015; Haddix, 2008; Terry & Irving, 2010). Though the core theoretical positions developed in this literature are not specific to any one type of linguistic diversity, much of the literature focuses on students who speak African American Vernacular English (AAVE).

African American Vernacular English (AAVE) is a dialect of American English largely spoken by members of the African American community. AAVE and Standard English (SE) vary along a number of dimensions, including having different grammatical rules; however, speakers of AAVE and SE rarely struggle to understand each other (Wheeler et al., 2012). Though AAVE and SE are linguistically on par—neither dialect is somehow objectively better than the other—people often think that AAVE is “not proper English” and speaking AAVE can mark one for discrimination in academic, professional, and social contexts (Baugh, 2002).

There are (roughly) two camps when it comes to thinking about linguistic diversity in the classroom: subtractive models and additive models. Additive models are additive because they aim to add a new behavioral repertoire (SE) to an existing behavioral repertoire, AAVE. By contrast, subtractive models are subtractive because they aim to introduce a new behavioral repertoire (SE) while eliminating (or subtracting) the preexisting behavioral repertoire, AAVE (Flores & Rosa, 2015). Subtractive models view the home language as a barrier to be overcome whereas additive models view the home language as an additional strength to be utilized (Flores & Rosa, 2015).

In rejecting the idea that SE is superior to other linguistic practices, additive models face a challenge in justifying instruction in SE. If SE and AAVE are equally valuable and students are already fluent in AAVE, why would it be important to offer instruction in SE?

In general, proponents of additive models justify instruction in SE in pragmatic terms. Though SE is not objectively better than AAVE, fluency in SE is de facto necessary to achieve socioeconomic success in the United States (Delpit, 2006). Given that there is discrimination against AAVE speakers, teaching students SE allows them to more successfully navigate a discriminatory world and gives them the tools they need to utilize “codes of power” (Delpit, 2006, pp. 24–25).

Code-switching is key to additive justifications for instruction in SE. Code-switching involves moving between linguistic practices to match the reinforcement available in one’s environment (Gardner-Chloros, 2009). Thus, someone may speak SE in school or at the workplace while speaking AAVE with friends or while at home. Code-switching makes good on the idea that SE is no better than AAVE (and vice versa); just as there are contexts in which SE will give superior access to reinforcement, there are contexts in which AAVE will give superior access to reinforcement. Teaching students to code-switch involves helping students learn about “the arbitrariness of those codes [of power] and about the power relationships they represent” (Delpit, 2006, p. 165).

Behavior analysts can take a similar approach to thinking about neurodiversity and behavior analytic interventions. Just as one need not think that SE is superior to AAVE in order to justify teaching SE, one need not think that neurotypical behaviors are superior in order to justify teaching them. Individuals who do not behave in a manner consistent with expectations of neurotypicality in certain situations can permissibly be taught (1) how to engage in contextually specific responses to better access reinforcement and avoid punishment; and (2) to discriminate (i.e., code-switch) the contexts in which it is advantageous to engage in one repertoire or another. Thus, within appropriate parameters and when done for the rights reasons, just as it can be permissible to teach SE, it can be permissible to teach contextually specific and pragmatically successful neurotypical behaviors.

Code-switching is essential to this conceptualization of behavior analysis. One can only hold that neurotypical and neurodivergent behavior are on par if there are contexts in which each is appropriate. If, in every context, neurotypical behavior is preferable, there is no sense in which neurodivergent behavior is equally good. Thus, just as teaching SE is permissible when it is paired with teaching students to discriminate between contexts in which various linguistic practices contact reinforcement, ABA can likewise teach autistic individuals to “code-switch” as appropriate.

Applying the Analogy to the Case Studies

This analogy between ABA and additive models of linguistic diversity does little to change the analysis of SELF-INJURIOUS BEHAVIOR, the case study in which Avery’s repetitive behavior was causing self-injury. Intervention remains justified by the need to prevent the objective harms of self-injury, even from the neurodiversity perspective.Footnote 9

Our original analysis of HAND-FLAPPING IN PRIVATE, the case in which Ace engaged in repetitive hand-flapping in the privacy of their own home and in a way that did not interfere with academic or social goals, highlighted the ways in which this request for intervention reflected the problematic view that neurodivergent behaviors are inherently bad or pathological. The analogy to additive models of linguistic diversity strengthens this criticism. Additive models highlight the ways in which negative attitudes toward linguistic diversity unjustly limit the opportunities available to those who do not speak SE. Likewise, negative attitudes toward neurodivergent behaviors like hand-flapping contribute to the discriminatory barriers faced by individuals with autism (Kapp et al., 2019).

HAND-FLAPPING IN PRIVATE serves to demonstrate the limits on the behavioral interventions that can be justified by an “additive” approach. Even with the perspective of an “additive” approach to intervention, some behavioral intervention targets will be impermissible. For example, in regard to the hypothetical case example of HAND-FLAPPING IN PRIVATE, it is unlikely that an intervention aimed at directly modifying this behavior (e.g., to reduce, eliminate, or replace it, by way of antecedent and consequence manipulations immediately surrounding this behavior) would be permissible.

The analogy to additive models of linguistic diversity does not, however, suggest that all interventions that might indirectly affect rates of HAND-FLAPPING IN PRIVATE are impermissible. Interventions designed to help Ace develop a broad range of play/leisure skills may lead Ace to contact reinforcement for other behaviors and build repertoires that may indirectly influence rates of hand-flapping.

Ace’s hand-flapping could also be a response to aversive environmental stimuli. Individuals with autism often have different reactions to sensory stimuli, relative to neurotypical individuals (APA, 2013), and repetitive motor mannerisms may be a response to exposure to aversive sensory stimuli that are negatively reinforcing as a form of self-regulation (Kapp et al., 2019; Zentall & Zentall, 1983). Modifying environments in order to make them more aligned with the needs of autistic individuals and their unique sensory responses could indirectly result in reduced rates of various types of repetitive behaviors, including hand flapping, by way of altering motivating operations (e.g., abating the motivating operation or negative reinforcement).

Although a reduction in hand-flapping may occur indirectly following certain environmental modifications or by teaching other skills, reducing Ace’s hand-flapping behavior should not be the goal of any such intervention. Designing interventions with the aim of reducing Ace’s hand-flapping appears to treat nonneurotypical behavior as if it were inherently bad or pathological.

The analogy to additive models of linguistic diversity yields the most notable results regarding HAND-FLAPPING IN SCHOOL. In this case, Alex’s repetitive hand-flapping in school is being considered as a target for intervention based on the grounds that it is interfering with Alex’s academic progress. In our original analysis, the abolitionist neurodiversity critique challenges the permissibility of this rationale, highlighting the ways in which the notion of academic progress tacitly treats the academic trajectory of neurotypical individuals as normative.

The analogy to additive models of linguistic diversity allows us to provide a competing understanding of the importance of academic progress. Additive models of linguistic diversity justify teaching SE by highlighting the ways in which fluency in SE serves as a key to accessing opportunities. The behavior analyst can take a similar approach in considering interventions for HAND-FLAPPING IN SCHOOL.

From an additive approach perspective, an intervention that targets hand-flapping in school is not justified on the grounds that hand-flapping is somehow inherently bad or wrong, nor is it justified on the grounds that neurotypical academic progress is normative. Rather, intervention may be justified on the grounds that it will give Alex the tools she needs to more successfully navigate varied contexts, including those in which neurotypical behavior has been historically favored. Intervention in HAND-FLAPPING IN SCHOOL is thus justified on the grounds that it gives Alex an increased repertoire of available responses that may help her maximize reinforcement across varied contexts.

As part of increasing client autonomy, behavior-analytic interventions may increase client independence, increase a client's indices of happiness (e.g., Green & Reid, 1996; Reid, 2016), increase a client’s range of preferences (e.g., Dutt et al., 2014; Winborn-Kemmerer et al., 2009), increase positive and reciprocal interactions between caregivers and their children (e.g., Ala'i-Rosales et al., 2013; Solomon et al., 2008), and decrease self-reported depression and anxiety (e.g., Hillier et al., 2011; Schiltz et al., 2018), to name a few examples. In each instance we recognize that there are goods that neurodivergent individuals struggle to achieve, not because they are neurodiverse, but because our institutional and social environments privilege neurotypicality. When individuals with autism are better able to choose between engaging in more and less paradigmatically neurotypical behaviors and repertoires across contexts, they widen the range of positive reinforcers available to them. In the words of Schuck, Tagavi, et al. (2022b), “[B]eing able to successfully wait in line will allow children access to a variety of environments and activities, and teaching this skill does not appear to be incongruent with the neurodiversity paradigm” (p. 4635).

Although we have shown that behavior analysts can offer interventions without devaluing autistic clients, it is important to stress the limitations of the analogy to additive models of linguistic diversity. As described earlier, the neurodiversity movement has raised a number of concerns regarding ABA. Some of these relate to the methods used in behavioral interventions (e.g. Devita-Raeburn, 2016) whereas others are concerned that masking neurodivergent behaviors is itself harmful (Cassidy et al., 2020; see also, Petrolini et al., 2023). The analogy to additive models of linguistic diversity speaks well to the abolitionist concerns raised by the neurodiversity critique but fails to address the broader range of reformist concerns. Though addressing reformist concerns raised by the neurodiversity movement falls outside the scope of this project, behavior analysis would be ethically remiss to ignore the sincere concerns of those it aims to serve. By considering and responding to the abolitionist neurodiversity critique, this article aims to expand the nascent discussions related to the neurodiversity perspective in our field, but there remains much work to be done to thoughtfully consider and respond to the range of perspectives and concerns of clients with autism.

Ethical Implications

We have now introduced the neurodiversity perspective, the abolitionist neurodiversity critique of ABA, and indicated how an analogy to additive models of linguistic diversity can help articulate at least some of the goals of behavior analytic interventions in a way compatible with the abolitionist neurodiversity critique. In this section we will highlight some of the concrete ethical implications entailed by the preceding discussion.

Reformist neurodiversity critiques of ABA aim to make the case that various reforms to the practice of ABA are morally required. Although this section also makes the case that taking the neurodiversity perspective seriously has ethical implications for the practice of behavior analysis, the following discussion continues to focus on the abolitionist neurodiversity critique of ABA. Although we have argued that the abolitionist neurodiversity critique can be answered via an analogy to additive models of linguistic diversity, embracing this analogy comes with concrete implications for the practice of behavior analysis. The goal of the following discussion is to highlight some of these implications.

Teaching Clients to Discriminate between Contexts

The idea that various linguistic practices are equally valuable is a centerpiece of additive models of linguistic diversity. Teaching code-switching makes good on this commitment; no language is uniquely privileged. Rather, some linguistic practices provide access to reinforcement in some environments, and other linguistic practices provide access to reinforcement in other environments (Neely et al., 2020).

In order to make good on the analogy to additive models of linguistic diversity, ABA interventions should include discrimination teaching/training to help clients identify contexts in which various behaviors are, and are not, likely to result in access to reinforcement. Reconsider HAND-FLAPPING IN SCHOOL. The target behavior—hand-flapping—is currently a barrier to achieving academic goals and may be (or become) a barrier to achieving social or professional goals. It may thus be appropriate to design an intervention to expand the repertoires of responses from which the client can choose.

Roughly speaking, there are two ways one could design such an intervention. One could aim for broad generalization, so that the frequency of repetitive motor mannerisms, including hand-flapping, decreases across all contexts. This approach suggests there is something inherently wrong or bad with repetitive behaviors associated with ASD, independent of the context in which it occurs.

By contrast, one could design an intervention that aims to help clients discriminate differential access to reinforcers for particular behaviors across contexts, which may result in differential rates of repetitive motor mannerisms (e.g., hand flapping) across contexts. For example, a client may be taught to recognize that engaging in certain repetitive motor mannerisms decreases the likelihood of accessing social positive reinforcers in some contexts (e.g., in academic or professional settings) and/or may increase the likelihood of contacting punishment in other contexts (e.g., during an encounter with law enforcement).Footnote 10 An ultimate goal should be to empower clients to choose between being in contexts in which hand-flapping may be more or less likely to contact various types of reinforcement (e.g., social vs. automatic reinforcement) and to adapt their behavior as appropriate (e.g., engage in an alternative, more socially acceptable response) in contexts in which punishment may be more likely.

Relative to targeting hand-flapping for reduction broadly, teaching discriminated responding and the ability to code-switch between contexts (Kingsdorf, 2014; Neely et al., 2020) better aligns with the neurodiversity perspective. This approach recognizes that there is nothing inherently problematic about repetitive behavior and that, in many contexts, there is no reason such behavior should be avoided. Clients are empowered to identify the contexts in which repetitive behaviors may prevent access to reinforcement (or encounter punishment) and are empowered to vary their behavioral repertoires to fit the contingencies of the environments in which they find themselves, as well as to choose between contexts and the reinforcers available in each.

The current Professional and Ethical Compliance Code for Behavior Analysts (henceforth, The Code) requires that, “Behavior analysts design behavior-change programs that are conceptually consistent with behavior analytic principles” (BACB, 2019, p. 12). As we argued at the outset, operant learning theory does not explain behaviors in terms of pathology, but rather as the lawful consequence of interactions between an organism and its environment (Skinner, Skinner, 1965, Skinner, 2019). As such, operant learning theory commits the behavior analyst to the view that autism and neurodivergent behaviors are mere-differences.

Aiming to decrease the frequency of repetitive motor mannerisms across all contexts suggests that neurodivergent behaviors are bad-differences and, as a result, is likely not “consistent with behavior analytic principles” (BACB, 2019, p. 12). By contrast, teaching clients to code-switch treats neurodivergent behaviors as mere-differences, rather than bad-differences, and thus remains consistent with both the ethical principles of The Code and the theoretical underpinnings of behavior analysis.

Likewise, The Code requires that, “behavior analysts do not engage in discrimination against individuals or groups based on . . . disability” (BACB, 2019, p. 5). Just as treating race or gender orientation as a bad-difference constitutes a form of discrimination, designing interventions that treat autism as a bad-difference is discriminatory. This reading of the code suggests targeting a nonharmful, neurodivergent behavior such as HAND-FLAPPING IN PRIVATE for reduction may not be permissible, and teaching code-switching in cases like HAND-FLAPPING IN SCHOOL may be permissible.Footnote 11

BCBA, Parent, and Teacher Training

Drawing on the analogy to additive models of linguistic diversity, we have suggested that behavior analysts should (1) hold that autism is a mere-difference; (2) emphasize the unique strengths of autistic individuals; and (3) highlight the importance of teaching individuals with autism to code-switch so that they can vary their behavior in response to changes in environmental contingencies. Though additive models of linguistic diversity represent an improvement over their subtractive counterparts, additive models have (reasonably) been criticized for being racist:

The basic assumption of bi-dialecticalism [i.e. additive models] is that the prejudices of middle-class whites cannot be changed but must be accepted and indeed enforced on lesser breeds. Upward mobility, it is assumed, is the end of education, but white power will deny upward mobility to speakers of black English [i.e., AAVE], who must therefore be made to talk white English [i.e., SE] in their contacts with the white world. (Sledd, 1969, p. 1309)

Likewise, teaching individuals with autism to code-switch seems to assume that prejudices against individuals with autism cannot be changed and that the burden of navigating this discrimination falls to individuals with autism.

Even those critical of additive approaches to linguistic diversity do not deny the importance of teaching students to code-switch. Rather, in addition they aim to shift “the conversation from trying to improve the linguistic practices of language-minoritized students toward challenging the ways that their linguistic practices are taken up and interpreted by the white listening subject” (Flores & Rosa, 2015, p. 167). Teaching speakers to code-switch cannot be the end of the story. Parallel efforts to educate listeners are required.

Designing behavior-analytic interventions as a type of code-switching is not, on its own, an adequate response to the abolitionist neurodiversity concern that the implementation of ABA interventions for patients with autism reinforces discriminatory social structures and practices. Merely teaching code-switching leaves broader discriminatory structures unchallenged and places the entire onus of navigating prejudice on those with autism.

Behavior analytic training programs should be redesigned so that “neurodiversity-informed pedagogy . . . [becomes] embedded into existing clinical programs to enhance the training of professionals and ensure they are capable of serving as the ‘neurodiversity advocate’ when working with Autistic individuals and their families” (Schuck, Tagavi, et al., 2022b, p. 4634). Furthermore, there is a deep and established literature that develops and evaluates protocols for training parents, teachers, and direct-care staff in ABA (e.g., Matson et al., 2009; Patterson et al., 2012). Parent- and teacher-education programs from outside of behavior analysis offer further valuable resources, e.g., SAFE (McKenzie et al., 2022) and SOLACE (Lodder et al., 2019). Alongside teaching clients to code-switch, parents, teachers, and direct care staff should be trained to recognize the ways that neurodivergent behaviors constitute a functional response to environmental contingencies and are fundamentally no different from neurotypical behavior.

The BACB Code of Ethics requires that, “behavior analysts do not engage in discrimination against individuals or groups based on . . . disability” (BACB, 2019, p. 5). On a narrow reading, this mandate merely requires that behavior analysts provide equal access to high quality ABA regardless of someone’s disability status. A broader reading is, however, also available. Behavior analysts that focus only on teaching clients to navigate ableist social and institutional environments reentrench the ableist assumption that society need not accommodate, much less celebrate, neurodiversity. Thus, The Code’s mandate that behavior analysts not engage in social discrimination can be read as a requirement to train caregivers and relevant parties, such as parents and teachers, to recognize that neurodivergent behavior is fundamentally no different from neurotypical behavior—each is a functional response to environmental contingencies.


The neurodiversity perspective and the medical model are each associated with a particular way of talking and writing about autism. The medical model tends toward terminology that treats autism as a pathology. The following terminology is thus generally associated with the medical model: “deficit,” “treatment,” “diagnosis,” “cure,” etc. Each of these terms suggests that there is something bad, wrong, or problematic about autism—to have a “deficit” is to be lacking in something whereas “treatment,” “diagnosis,” and “cure” all imply that there is an underlying pathology to be treated, diagnosed, or cured.Footnote 12

Behavior analysts should be careful about the language they use when discussing clients. Avoiding terminology associated with the medical model best fits with the behavior analyst’s theoretical commitments while also challenging predominant ableist assumptions (Schuck, Tagavi, et al., 2022b). Furthermore, families who reject the medical model are likely to find the use of associated terminology off-putting (Nicolaidis, 2012). Building rapport with families may thus require care in behavior analysts’ linguistic choices, as well as explicitly communicating with families about their linguistic preferences. Happily, a number of guides to nonpathologizing language are now available (for further discussion and a list of valuable resources, see Schuck et al. [Schuck, Tagavi, et al., 2022b, p. 4634]).

Just as The Code mandates that behavior change programs be conceptually consistent with behavior-analytic principles, it likewise requires that behavior analysts “use language that is . . . conceptually systematic with the profession of behavior analysis” (BACB, 2019, p. 5). As such, The Code’s guidance regarding behavior analysts’ use of language may forbid the normatively laden language of the medical model.

Attempts to avoid terminology associated with the medical model are complicated by the fact that the language of the medical model is interwoven throughout the medical establishment.Footnote 13 Thus, communicating with other health-care providers and perhaps with insurance companies may evoke the use of pathologizing terminology. Given that contingencies may differentially reinforce the use of pathologizing terminology, behavior analysts should consider discussing these contingencies with their clients and/or their caregivers. The use of pathologizing terminology in specific contexts could be understood as another instance of code-switching; in order to access reinforcement across diverse contexts, behavior analysts may choose to vary their behavioral repertoires.Footnote 14

Incorporating Preferences

Autism is defined in part by “restricted, repetitive patterns of behavior, interests, or activities,” including “highly restricted, fixated interests” as well as “unusual interests in sensory aspects of the environment” (APA, 2013, p. 50). Thus, individuals with autism often have strong and/or idiosyncratic preferences. Taking neurodiversity seriously requires respecting the preferences of individuals with autism.

Incorporating preferences into intervention, including idiosyncratic and/or repetitive interests, expresses a recognition that the preferences of individuals with autism are valuable, worthy of respect, and often constitute a unique strength generally not shared by neurotypical peers. There is a wide range of empirically validated preference assessments available for use across a comprehensive range of demographics and settings (DeLeon & Iwata, 1996; Dutt et al., 2014; Fisher et al., 1992; Lill et al., 2021; Pace et al., 1985; Resetar & Noell, 2008; Roane et al., 1998; Simonian et al., 2020; Virués-Ortega et al., 2014), and for a variety of types of preferences, including different types of social interaction (e.g., Butler & Graff, 2021; Conine & Vollmer, 2019; Morris & Vollmer, 2020; Nuernberger et al., 2012). Behavior analysts should be highly proficient in utilizing preference assessment methodologies, frequently assess clients’ preferences, and incorporate these preferences into treatment whenever possible.

Future Directions: The Appropriate Aims of the Practice of Behavior Analysis

In addition to having implications at the clinical level, the abolitionist neurodiversity critique highlights an important gap within ethics discussion in ABA. The current version of The Code has a wide range of guidance regarding how ABA-based interventions should be practiced—ranging from the recruitment of research participants to the storage of records. It is, however, largely silent on what therapeutic goals are appropriate targets for intervention. Thus, though The Code is clear that “clients have a right to effective treatment” (BACB, 2019, p. 8), there is no substantive engagement with the question of the ethical boundaries that constrain the aims of that treatment.

This problem is not unique to ethical considerations regarding behavioral intervention and autism. The problem of the aims of intervention highlighted by the abolitionist neurodiversity critique are perhaps easiest to understand in the context of conversion therapy, a behavioral intervention that is widely understood to be impermissible by the behavior analytic community (ABAI, 2021) yet is often treated as morally indistinguishable from ABA intervention with autistic clients by members of the neurodiversity movement (Autistic Collaboration, n.d.).

By way of example, consider the case that DeFelice and Diller (2019) make against the permissibility of conversion therapy. DeFelice and Diller note that, though some behavior analysts historically engaged in research aimed at changing individuals’ sexual orientation, much of this research “no longer meet[s] the ethical requirements for our field” (p. 3). The research in question falls short on a number of dimensions, including its emphasis on punishment procedures, failures in informed consent, failure to get appropriate review by ethics boards, and failure to maintain the “dignity and welfare of research participants” (DeFelice & Diller, 2019, p. 3).

DeFelice and Diller’s (DeFelice & Diller, 2019) analysis of the impermissibility of conversion therapy is accurate, but it may also be incomplete. Even if the ethical shortcomings highlighted by DeFelice and Diller were addressed, behavior analytic research into so-called “conversion therapy” would remain ethically problematic. Independent of its implementation, conversion therapy treats homosexuality as a problem to be solved, rather than as a difference to be celebrated. No analysis of the ethical shortcomings of conversion therapy can be complete without highlighting the fact that the aim of conversation therapy is, by itself, problematic.

Once we have recognized the structure of the problem in the ethically unambiguous context of conversion therapy, it is easy to spot in the more controversial context of the ethics of ABA intervention for individuals with autism. The abolitionist neurodiversity critique highlights this concern: independent of its behavioral implementation, the aim of teaching autistic individuals to behave as if they were neurotypical may, itself, be problematic (see, e.g., Schuck, Tagavi, et al., 2022b). This is not a criticism of the way in which ABA is implemented, but rather of the potential goals of intervention.

The newest version of The Code, which became effective in 2022, makes important progress in considering the appropriate aims of ABA. For example, the new version of The Code requires that behavior analysts “maximize benefits and do no harm” (BACB, 2020, pg. 4). Though the current Ethics Code for Behavior Analysts offers a substantial improvement over its predecessor, this discussion of the aims of ABA remains too limited to answer the concern raised by the abolitionist neurodiversity critique. Sexual orientation is a mere-difference. It is neither better nor worse to be heterosexual (or gay). Behavioral interventions that aim at reducing gay behavior are thus wrong, in part, because they cause harm (DeFelice & Diller, 2019), but not only because they cause harm. Likewise, because neurotypicality and neurodiversity are mere-differences, an adequate analysis of the wrongness of interventions that aim at neurotypicality cannot be understood solely in terms of harms and benefits. Some further theoretical apparatus is required to make sense of the wrongness of such interventions.

Social validity may appear to hold promise as the needed theoretical apparatus. In his seminal article, Wolf (1978) identifies three components of social validity: “the social significance of the goals,” “the social appropriateness of the procedures,” and “the social importance of the effects” (p. 207). Social validity speaks directly to the selection of treatment goals and may thus seem to provide the necessary constraints on the aims of behavior analytic intervention. Furthermore, by fostering a collaborative relationship between researchers and participants, social validity serves as “a safeguard to promote inclusion of communities, include the voices of members, and prevent exploitation and further marginalization of persons with vulnerabilities that participate in research” (Pritchett et al., 2021, p. 4).

Despite the importance of social validity, members of the neurodiversity movement have identified ABA’s emphasis on “social significance” as contributing to ableist bias in ABA (Chapman & Bovell, 2022). If one accepts that bias against neurodivergent behavior is widespread, setting therapeutic goals that align with neurotypical parents’ priorities may serve as a vehicle for entrenching bias in behavior-analytic intervention, rather than serving as a way to combat such bias.Footnote 15 At a more general level, although social validity does constrain the goals of behavior analytic intervention, it does so in a way that often reflects the dominant societal perspective. The problem is that dominant societal perspectives are often racist, sexist, homophobic, and ableist (Clair & Denis, 2015; Homan, 2019; Nario-Redmond, 2019; Rinaldi, 2020).

The challenge of finding the right theoretical apparatus for constraining the goals of ABA is likely to be uniquely difficult for behavior analysis. Consider, by way of contrast, the standard answer given about the appropriate aims of medicine. In brief, the standard account holds that medicine should aim at addressing suffering (in its varied forms) caused by pathology (Pellegrino, 1981). The nature of medicine—the fact that medicine seeks to address pathology—constrains the goals at which it is appropriate for medicine to aim while nonetheless leaving wide latitude for individual differences and preferences.

By contrast, ABA’s great strength is that it can be applied to nearly every behavior, whether or not the goals of the intervention are morally appropriate. ABA can thus be used to increase recycling behavior (e.g., Ludwig et al., 1998), to try and reduce noncisgender behavior (e.g., Rekers & Lovaas, 1974), to reduce self-injurious behavior (e.g., Vollmer et al., 1993), or to try and get individuals with autism to act as if they were neurotypical. Yet, as the abolitionist neurodiversity critique makes clear, behavior analysis owes some account of the boundaries of appropriate goals for intervention. Important work remains to be done in identifying what behavioral goals are, and are not, appropriate for behavior-analytic interventions.


Though the neurodiversity movement includes the voices of a wide range of advocates, the behavior-analytic literature on neurodiversity remains limited. This article aims to begin filling in this lacuna by reviewing the conceptual foundations of the neurodiversity perspective, by showing that the neurodiversity perspective is in line with core behavior analytic commitments, by reviewing and critically evaluating the abolitionist neurodiversity critique, and by considering the broader implications for both practical and theoretical ethics in ABA. This is a small—but we hope an important—step in understanding how ABA can be improved by listening to the voices of its clients with autism.