22/08/2022
Why ABA keeps failing, and fails to see its own failure
[CW/TW: trauma, torture, child abuse, cults, coercion, mention of genocide, basically everything bar clowns]
[ID: inverse colour (golden yellows flipped to blues) closeup of solar storming on our sun. Text in Roboto typeface overlays the image in a very pale blue. Text reads: "Why ABA keeps failing and fails to see it's own failure"]
[Yes the colour choices and use of Roboto are symbolic. Well done! You can have a skittle.]
Applied Behavioural Analysis (ABA) as an approach to addressing psychological issues is notable for its proponent’s incessant self-congratulation on stellar success, on a level that borders on propaganda.
By contrast, those subjected to ABA are close to unanimous in viewing it as not just failing to achieve its goals, but actually harmful, a subjective observation backed up by a growing body of studies into the long-term negative effects of ABA, and serious questions about both the ethical basis for ABA and the validity of its claimed successes.
There is no doubt that views are strongly polarised. So who is right? Is either ‘side’ right? Is there a ‘good ABA’ to be had? Can you mix some ABA techniques and methods with other stuff and not do harm? Is everything that deals with ‘behaviour’ inherently problematic?
Let’s start by going back a century or so…
Psychology was in its infancy, biology in its pre-teens, we had no real clue about DNA and genetics, our understanding of things like trauma and developmental psychology was effectively non-existent, there were no brain scans or imaging. Eugenics was considered as a good idea by many (too many), as was fascism, racial segregation, most women had no vote, most children got no second level education. Definitely no WhatsApp or TikTok. It was most certainly a very different world.
Behavioural theory was established as an important part of psychology, and ideas like ‘a positive outcome tends to increase a behaviour’s frequency, a negative outcome tends to decrease its frequency’ (Law of Effect) were demonstrated, well understood, and part of established thinking. John B. Watson argued that objective analysis of ‘the mind’ was impossible and thus should be ignored, and that basically everything (barring [ahem] ‘trivialities’ like society and language) could be analysed in terms of the interaction of learned behaviour and the inherent qualities of an organism. You could, he argued, manipulate behaviour by designing what and how an organism learned, and that all organisms essentially learned in the same way, even though they may vary in the rate of learning. What you learn by experimenting on rats is applicable, then, to cats or monkeys or humans.
There’s an awful lot going on there, and I’m guessing you will be dubious about at least some of it, and for good reason. To start with, pushing aside the vast influence of words and their associated complex webs of tone, meaning, nuance, metaphor, context when dealing with humans is… naïve to say the least. Doing the same with the closely linked complex webs of social activity, influence, precedent, hierarchy, sense of identity, sense of entitlement, rights, opportunity, and more is not naïve, it is shockingly foolish. If I told you that influencers, celebrities, sports stars, peer pressure, workmates, bullying, social climbing, populism, fads, fashions, gaslighting, coercion, social class, ethnicity, gender norms, and discrimination might be able to alter a person’s behaviour, but it is okay to ignore them and you can manipulate an individual’s behaviour without taking them into account, you would rightly consider my opinions meaningless, even deliberate trolling.
Equally, pretending that fears, hopes, aspirations, imagination, beliefs and other internal factors can be ignored because they cannot be objectively observed is just silly. Reckless, even. Similarly for the assumption that transferring techniques between species is straightforward. The question of the ethics of deliberately manipulating a person’s behaviour through simplistic punishments and rewards we have not even touched on. The ethical issues hardly need detailing.
In response, B.F. Skinner highlighted several of the issues noted above and produced a much more nuanced and balanced perspective in the mid 1940s. Unfortunately he also developed a method called operant conditioning (shaping a behaviour by associating it with a separate reinforcing experience, either positive or negative) in which he declared that all nuance and balance could be ignored. Essentially, ‘Watson was wrong in the following ways, but let’s do things as he suggested anyway.’
It seems almost incredible when it is written like that but that is essentially what happened, giving us what we call radical behaviourism as a philosophy, and applied behavioural analysis as its application.
These are not chancers peddling self help books or personal improvement videos, they are key figures in the history and development of psychology who have had considerable influence across many areas of human society in the decades since. This paints a very negative picture of these two men and by extension of psychology which is unfair but for our purposes here (considering the what, why and how of ABA as used on Autistic minors) the perspective is valid.
Every extremism needs a zealot and a dread enemy, and ABA thus has Ole Ivar Lovaas as the former, and Autistic children’s behaviour as the latter. Lovaas discovered behaviourism in the late 1950s and established his now infamous Young Autism Project in 1962 at UCLA, using operant conditioning as a technique to achieve behaviour modification. Inevitably, part of the package was treating everything bar the stimulus and behaviour as a ‘black box’ and thus irrelevant – all the culture, aspirations, memories, affections and so on Watson and Skinner assured us was unknowable.
At this point the importance of ethical considerations should be paramount. He was working on minors without independent oversight, without their consent (and often with disregard to their protests and distress), and using methods and theory that disregarded everything even vaguely human about his subjects. The results were a pursuit of ‘errorless learning’ (repeated perfect ex*****on of a given task) by way of slapping, abusive shouting, restraint, constant mindless repetitive tasks, excessively long duration sessions, no option for the subject to time out or escape, and even the use of electric shock to punish errors.
During a wartime scenario, treating a prisoner in this way would be classed as torture and a war crime; in any circumstance this would count as a human rights abuse; being, as it was, directed at a clearly defined, vulnerable, effectively imprisoned, minority of society and not used on other social groups, this is a crime against humanity and if the definition of genocide did not specify that the agent must be a political state, there would be grounds to consider this activity under that heading also.
Considering the clinical ethics, with this in mind, is moot. That he was not arrested and charged is disappointing. That he was allowed to continue in his role as an educator and experimental psychologist is shocking. Literally shocking in the case of Lovaas.
He called his technique Discrete Trial Training (DTT), where a child was typically given an instruction, then a prompt to the ‘correct’ response, the child responds (correctly or not), and followed by a stimulus reinforcer. In Lovaas’ case, ‘stimulus reinforcer’ here means electric shock, yelling in a child’s face, slapping, etc. Each session continued for hours and hours. DTT is the key method used by ABA practitioners to this day to carry out operant conditioning on Autistic minors. In most (sadly, not all) instances, the reinforcer is ‘nicer.’ The psychological component, the infantilisation, the vulnerability, the relentlessness and the trauma are unchanged even if kids in most places today are not actually given electric shocks.
Yikes, right?
And yes, ‘yikes’ is me being overly polite for effect. Say the words in your head out loud – think of it as audience participation.
Oh, but… ABA is ‘evidence based.’ It says so on the box. Several times, actually.
Almost all studies on ABA have been flawed, thus adding an additional layer of questions about whether the approach even achieves the level of success it claims. Study groups have generally been small, or just a single subject, almost none have a control group, those that do rarely have randomised assignment of subjects to groups, the categories and measurement of outcomes focus on simplistic elements of ‘development’ or arbitrary ‘tasks,’ often positive results rely on the opinion of the subjects’ parents and thus are inherently biased through being invested in successful outcomes, results take little or no account of expected or actual progress in the subjects’ peer groups, rob subjects of excessively large periods of time from their normal childhood, and rely on coerced engagement, thus making any results skewed. Dangerously skewed. None of this counts as good science. Much of it has nothing to do with science. 50 years ago it was, at best, bad science, and it is all the more questionable today.
In essence, when we look at research on the long-term effects of ABA, the transferability of any gained skills or capability, the words of those who have experienced it, or any human rights or ethical concerns, all are near unanimous in condemning ABA as deeply and consistently harmful and lacking in any significant tangible benefits that cannot be more effectively and less harmfully achieved through speech and language, occupational or physical therapy – or indeed without use of any therapy intervention at all.
But even before looking at these we should actually expect ABA to be a failure. It should not achieve its aims, and we should expect harm would come to those subjected to it. We should expect it to fail simply because the theoretical basis for ABA is massively flawed and based on simplistic – even archaic – ideas and knowledge, and the methods used are degrading and dehumanising. We know it should do harm. And guess what? It does. Everything in the previous paragraph shows the harm it does.
How, then, can ABA practitioners declare their work as “evidence based”?
The answer relies in part on basic propaganda techniques – tell people a lie often enough and it becomes truth – but also, when pushed, appeals to authority such as university programmes, research papers, text books, college professors and professional bodies, all of which themselves are applying the same bias and poor science and incessant campaigning in endless mutually-referencing loops, all built on the same highly questionable theoretical foundations and the assertion – regularly heard – that those who disagree are wrong because they, frankly, are too damn thick to understand.
No science is built on the combination of bluster and arrogance, bias and marketing. That’s not science, it is, at best, aggressive product marketing and at worst cultish propaganda.
Behind all this is one word, though: arrogance. A refusal to accept that something might be wrong, that we, highly trained professionals, might have got it wrong, and terribly, destructively so. The ego can be a powerful master, especially when one’s entire career and purpose in life is founded on a theory that discounts it as irrelevant. You cannot see the error when you refuse to even look. And thus the ABA world both fails and refuses to admit failure is possible.
There was a time when we were comfortable talking about fellow humans as having an affinity for something, or a talent, or being unsuited to a particular activity, of being introspective or thoughtful or ingenious, a novel thinker, imaginative, enjoying their own company or the company of like-minded souls, of being a curious fellow or a quirky lass. In casting aside the fuzzy worldview of the past we have gained enormously by eliminating notions of brats and bold children, idiots and imbeciles, changelings and possession but also lost so much by way of accommodation and curiosity, diversity and opportunity. And dignity, respect, patience, compassion, consideration and the sheer delight and creativity that the endless diversity of humanity brings to our lives.
We are not machines, nor rocks. None of us. We are each and every one human, whole, complete, complex, changing and growing, The same but different. Different yet the same.
Diversity is the human normal.
And ABA… ABA is a terrible error. But it is an error we have it within our power to remedy.
Here is how:
Click on this link and add your signature.
https://my.uplift.ie/petitions/abolish-conversion-practices-on-neurodivergent-people
