The Aspergian

Is ABA Really “Dog Training for Children”?  A Professional Dog Trainer Weighs In.

ABA Therapists like to talk shop with me when they find out that I’m a dog trainer.

“I use the same principles in my work!”  they always say eagerly.  “It’s all operant conditioning, isn’t it?”

“Well,” I say, “a lot of it anyway.”

I love operant conditioning, and as a dog trainer, I’ve built my career around it.

I hold a Bachelor of Science in Psychology and I took every course on behaviour modification and conditioning that my university offered.  I use it daily when training assistance dogs.

So you can imagine I was curious about ABA when I heard about it.

I started reading ABA websites, the principles and goals involved.  The more I looked into ABA, the more I was exposed to the controversy around it.  While parents whose children have received ABA sing its praises and describe it as the therapy that saved their child, the adult autistic community seems to feel differently.

I discovered that autistic adults consider it abusive, and many who were subjected to it as children claim to have been emotionally damaged.

Some preliminary studies even suggest that adults who received ABA as children are at an increased risk of suicide and PTSD.

And quite commonly on Twitter, I’ve seen people call ABA “dog training for children.”

When I see that, I tend to go on Twitter rants in reply to it, because from everything I have read and seen of ABA, it is NOT “dog training” for children.

…I would never treat a dog that way.

What is ABA?

I’ll let ABA proponents explain it themselves:

Autism Speaks says:

ABA therapy applies our understanding of how behavior works to real situations.  The goal is to increase behaviors that are helpful and decrease behaviors that are harmful or affect learning.

Autism Canada says:

Behaviour Analysis is the science of behaviour.  Applied behaviour analysis (ABA) is the process of systematically applying interventions, based upon the principles of learning theory, to improve socially significant behaviours to a meaningful degree.

The Applied Behaviour Analysis (ABA) approach teaches social, motor, and verbal behaviours, as well as reasoning skills.  ABA treatment is especially useful in teaching behaviours to children with autism who may otherwise not “pick up” these behaviours on their own, as other children would.

AppliedBehaviourAnalysisEdu.org says:

Applied Behavior Analysis (ABA) is the practice of applying the psychological principles of learning theory in a systematic way to alter behavior in humans or animals.  The practice is used extensively in education, healthcare, animal training, and business management.  It is particularly prominent in the treatment of Autism Spectrum Disorder (ASD), for which it is one of the only scientifically valid therapies available.

Right away a few things in these descriptions strike me as odd. 

For one thing, none of them—or any of the information I’ve found from the ABA discipline–make any reference to the autistic person’s emotions or well being.

It’s all about increasing certain behaviours and decreasing other behaviours, as if their student were a passive recipient. 

Dog trainers don’t talk about systematically altering behaviour as if the dog weren’t a thinking, feeling, sentient being. 

Look at Karen Pryor, who changed the face of dog training by introducing behaviourism to the wider world.  If you’ve ever heard of “clicker training”, you can thank Karen Pryor.

Clicker training is an animal training method based on behavioral psychology that relies on marking desirable behavior and rewarding it.  Desirable behavior is usually marked by using a “clicker”: a mechanical device that makes a short, distinct “click” sound which tells the animal exactly when they’re doing the right thing.  This clear form of communication, combined with positive reinforcement, is an effective, safe, and humane way to teach any animal any behavior that it is physically and mentally capable of doing.

That’s a pretty scientific way to describe dog training, isn’t it?  It sounds a lot like the description of ABA.

But notice this difference – this description uses words like “communication,” “safe,” and “humane.”

Even this brief description of operant conditioning as it is used on dogs addresses the dog’s point of view (“tells the animal exactly when they are doing the right thing”) and addresses the emotional and physical well being of the dog.

That’s a big difference.  The language of the ABA research and its guiding documents struck me as excessively clinical and distant, especially considering that they were discussing not animals, but human children.

Behaviour Analysis Is Not Dog Training

As a dog trainer, what also stood out for me when learning about ABA was the frankly-inaccurate claims that ABA is used on animals.

That’s just not correct. I have a science degree in behavioural psychology and I never heard the term “behaviour analysis” at university.

If you google “behaviour analysis,” you will only find sites that refer to autism and intellectual disability.  It simply does not exist outside of the realm of “ABA.” So their claims that they took an existing science and applied it to autism is extremely misleading.

It is clear that ABA employs some aspects of B.F.  Skinner’s radical behaviourism and this is what they mean when they say “behaviour analysis.”

Not only is it recognizable to anyone familiar with it, but you can also confirm the connection by googling radical behaviourism.  Everything will tell you it underpins ABA.

The founder of ABA as it exists today, Ivar Lovaas, who is also the father of gay conversion therapy, derived the principles of his therapies from radical behaviourism.

What is Radical Behaviourism? 

It’s hard to make this a short answer because the real answer involves a lot of bickering about minutiae among behaviourists, but the “tl;dr” of it all is that radical behaviourism believes that everything we do is a behaviour.

  • Your thoughts are behaviours.
  • Your feelings are behaviours.
  • All of them can be modified or altered through reward and punishment as consequences.

Radical Behaviourism is considered out-of-date by modern psychologists. 

While its principles do work and have largely been upheld in experimental research, cognitive science has found that there are some things that it can’t explain.

Language, for example.

B.F.  Skinner tried to explain language using behaviourism, but there is a lot in psycholinguistics that frankly cannot be explained through behaviourism.

Some things are larger than reward and punishment.  Empathy, for example.  Creative language.  Storytelling.  Music.

So basically, Radical Behaviourism is broadly seen by psychology professionals as a simplistic and restrictive theory which is useful in certain situations but cannot sum up the entirety of the human experience.  It doesn’t even satisfactorily answer some questions about behaviours seen in animals.

I can understand why Applied Behaviour Analysts decided to rename it as “Behaviour analysis,” but a rose by any other name is still radical behaviourism.

In any case, very few dog trainers use the radical behaviourism that’s employed in ABA.

Most of the dog trainers I know mix and match behaviourism with other cognitive science research and other methods to create a more holistic approach to training their dogs.  This is because dog trainers understand the limits of behaviourism on canines, because it doesn’t address the whole dog.

One would hope that someone considering using radical behaviourism on a human being would also recognize its limits.

So if it isn’t sufficient to properly train a dog, is it sufficient in educating a child?  Let’s take a look at ABA and the success rate that ABA proponents often boast about.

Does Radical Behaviourism Work?  How is “Success” Measured?

ABA is good at accomplishing what it set out to accomplish.  There is plenty of research demonstrating this, and ABA practitioners point to it often.  This doesn’t surprise me.  The principles behind radical behaviourism are sound and well proven.  You can use it to change an organism’s behaviour.

But should you?

Dog trainers spend a lot of time debating what is moral or ethical when it comes to changing the behaviour of their canine students.

They discuss not only what behaviours should be taught and which are not, but also what methods are ethical and which are not.

Do ABA practitioners do the same for the children in their care?

Are their goals in alignment with the best interests of the child?

Are their methods of achieving those goals humane and considerate of the child’s emotional and mental health?

Autism Speaks says that ABA helps:

· Increase language and communication skills
·  Improve attention, focus, social skills, memory, and academics
· Decrease problem behaviors

Autism Ontario says that:

An ABA program should address the core features and characteristics of ASD (i.e.  social skills, communication, and repetitive patterns of behaviour), as well as any barriers to learning (i.e.  challenging behaviours, stereotypy, etc.)

What struck me about these lists was how outwardly focused they were.  They really are teaching behaviours, aren’t they?

Nothing about improving the child’s quality of life, relieving anxiety, or helping the child cope with sensory sensitivities.

ABA is focused on shaping an autistic child to behave more like a non-autistic child, even to the point of shaping the child to play more like a non-autistic child.

Those are weird goals, quite frankly. 

None of these goals refer to improving the quality of life of the child.

I suppose that to some it would be redundant.  People seem to think that if you act normal, then you must feel normal.  But that’s not true for people.

It’s not even true for dogs.

The Problem of Operant Conditioning in Dog Training

In my field of dog training, you can use reward and punishment to train a dog not to bite when another dog passes by, or when a human touches it.  In fact, Cesar Millan does it with great aplomb on TV.

While he talks scientifically disproven nonsense about dominance, what he is really using is operant conditioning, the aspect of behaviourism which ABA therapists employ.  By punishing “problem behaviours” Cesar makes the dogs stop growling or attacking.

It looks like magic to viewers.

But to certified dog trainers like me, it looks like painting over rotten wood.  Worse.  It looks like burying a land mine.

We all know that we can feel angry without expressing anger.  That we can smile when inside we are crying.  You can stop someone from expressing an emotion, but that doesn’t make the emotion go away.

A dog who has been trained not to growl is considered by trainers to be a “time bomb dog.”

When you read about a dog attack that came “out of nowhere” and “without warning,” it is because this sort of method was used to handle “problem behaviours.”

Studies show that dogs trained with these sorts of methods actually have an increased rate of aggression, because punishing aggressive behaviour doesn’t deal with the underlying fear and anxiety that caused the aggression in the first place.

Certified dog trainers help the dog get over their fear…  and that uses no operant conditioning whatsoever.

Deal With The Underlying Problems Before Tackling Behaviour

Before a dog trainer breaks out the operant conditioning, our first task is to ensure that all of the dog’s fundamental needs are being met. 

Dogs need to run.  Dogs need to dig.  Dogs need to chase things.  Dogs need to chew things.  Dogs need to play with other dogs.  These are fundamental needs that cannot be removed from a dog’s psyche, and that can result in a lot of “problem behaviours” if they are not met.

If the dog is afraid or fearful, we gently encourage them to have positive experiences with the things that frighten them.  It seems counterintuitive to learn that feeding a dog who is growling at you will end the growling but it does – when the fear is gone, so is the aggression.  No operant conditioning required.

There are many times when an owner wants me to “train” the dog out of a behaviour and I refuse because either the behaviour is fear-based, or because the goal of the owner is unreasonable/unethical.

Before you train a dog you need to accept that it is a dog.

Yes, I’m happy to use operant conditioning to teach your dog to greet people politely at the door or to pull your wheelchair (which they love doing anyway) or to lie quietly on their bed during dinner. 

But when people ask me to train their retrievers not to pick things up in their mouth, or their terrier to stop digging, or their puppy to stop mouthing them, I have to sit down with them and have a little talk with them about the fact that they have a dog.

Could I train a retriever not to pick things up?  Yes.  If I broke out a shock collar I could put an end to it really fast.

But I won’t do it.

The fact is that you bought a retriever and holding things in their mouths is what they love.  You bought a terrier and they love to dig.  You bought a puppy and they love to play. 


Relieve fear instead of training the dog not to show its fear.  Teach your retriever to retrieve and buy a sandbox for your terrier instead of forcing your dog to ignore fundamental instincts. 

It’s not only easier – but it’s much kinder.

How Does ABA Ignore Fundamental Needs in Children?

One of the reasons parents and ABA professionals get so upset when autistic people call ABA “abusive” is the fact that they care deeply.  They genuinely want to improve these children’s lives.  Yet the vast majority of autistic people when polled (typically 97%) oppose ABA including and especially those who went through it as children.

Why is there such a disconnect?

Some of it has to do with a breakdown in the way autism is perceived.  Non-autistic people believe that “normalcy” is a fundamental need; indeed, a stated goal of ABA is to make the autistic child “indistinguishable from [neurotypical] peers.”

They think a child who blends into the crowd is a happy child.

When parents see their child engaging in unusual behaviours such as flapping, or ignoring other children, they see a child who is ill or damaged. 

When they see that child talking and working well at their desk and playing with other children, they see a child who has been healed.  Helped.  Saved.

If only they would listen to the autistic adults who are trying to tell them that this is not necessarily the case.  Because in reality, a happy autistic person may not look neurotypical. 

Knowing how to stack blocks or how to suppress essential means of regulation and expression (such as flapping) doesn’t make an autistic person “happier.” Often, in fact, it makes them less happy.

When you are autistic, talking can be exhausting.  Even if you are extremely verbal like I am – adults praised my precocious vocabulary as a child, and I have often been called a “chatterbox” – vocal speech is draining.  My well of words may be deeper than that of most other autistic people, but it is not bottomless.

Therefore, an autistic child who has been playing, without speaking much, all day may be a more rested and happier child at the end of the day than a child who has been chatting up a storm. 

A child who is given AAC or other ways to communicate their needs will be more comfortable and better regulated than a child who has been required to speak aloud… assuming they are even able to meet this demand.  An autistic child who spins or jumps is probably stimming and/or self-regulating through movement—ways to stay calm. 

Flapping and echolalia (repeating words or phrases), similarly, are expressions and often play an important emotional role as well as a developmental role.  Echolalic speech helps autistics, many of whom process language in a different part of the brain, to process the language they have heard and understand the meaning of the words.

Yet, ABA seeks to “extinguish” these things.

A good dog trainer doesn’t extinguish behaviours which improve the dog’s mental health and happiness.  But an ABA practitioner may not think twice before doing this to a human child.

In my field, when a dog loves to tug, I don’t operantly condition the dog to play tug less frequently.  Instead, I play more tug and incorporate obedience routines with it so that the dog has a blast learning to sit and stay.  And, if a dog has problems sitting still, I see that as a sign that the dog needs to move.  Some dogs are not meant to lie still.  Some dogs are meant to play agility and flyball, not lie on the couch all day. 

Sometimes I have to tell an owner that their dog’s personality just isn’t suited to what they are trying to achieve.

Could I achieve it?  Yes. 

I could break out the hardcore reward/punishment and using a combination of learned helplessness and positive reinforcement.  I could change that dog.  I could break their spirit.  Crush their ability to respond to their own emotions.

But I won’t do it.   Because the methods I would have to use would violate the Code of Ethics I agreed to when I applied for my certification through the Certification Council of Professional Dog Trainers.  The methods would violate the recommendations of every veterinary medical association.  The methods would violate the standards of the Society for the Prevention of Cruelty to Animals.

Some goals just aren’t worth it.

…But as it turns out, the Behaviour Analyst Certification Board has no such guiding principles.

Codes of Ethics: Dog Training versus ABA

Like all certifying bodies, the Behavior Analyst Certification Board has a professional code of ethics which its certificants must abide by to remain in good standing.

You can read them here.

I was amazed when I read it, because while it goes into great detail regarding what certificants may do with regard to the business aspect of things, right down to detailed guidelines on what you can do and say in the media, there is virtually nothing about the welfare of the therapy’s recipients.

The Certification Council of Professional Dog Trainers (CCPDT), by contrast, dedicates almost the entirety of its code of conduct to what, when, and how you interact with the animal.  While it also covers ethical business practices, its primary concern is the well-being of the “learner”.

First of all, the CCPDT has an entire policy dedicated to specifying in great detail what sorts of methods are condoned and which are not.  The Certification Council of Professional Dog Trainers policy is called “Least Intrusive Minimally Aversive” and is referred to in the code of ethics by its acronym, LIMA. 

This policy also lays out the “Humane Hierarchy” which advises which methods to start with first (ensure basic physical and emotional needs, then redesign the environment to remove antecedents, then employ positive reinforcement, then use differential reinforcement of an alternative behaviour, and so on), and which methods are only acceptable as a last resort.

The CCPDT Code of Ethics refers frequently to this policy which forms the backbone to the board’s entire framework.

If one applied the dog trainer’s code of ethics when working with an autistic child, therefore, one would first ensure that the child’s basic needs were met.  If the child seemed to be under-exercised or lacked sufficient playtime for an average child’s needs, they would recommend trying that first.

If that failed to make an impact, one would then pay attention to what situations impact the child, such as loud and noisy public places.  Following the dog trainer’s code of conduct, one would then recommend reducing exposure to these places, or providing the child with ear defenders and sunglasses, or gentle exposure based on the feedback of the child. 

One could not even consider other methods until these had first been tried and found unsuccessful.  For example, counter-conditioning or desensitization could only be considered if one could not simply keep the child away from the situation and if ear defenders, sunglasses, etc.  were not proving helpful.

One certainly could not deprive the child of playtime past the point needed by neurotypical children as part of any training plan, because that would violate one of the first steps in the Humane Hierarchy.

This step-by-step guide provided in the CCPDT’s Code of Ethics is designed to ensure the dog’s emotional and physical needs are considered first and foremost and to provide a clear guide of acceptable training conduct to its members.

The Well-being of Humans, The Well-being of Dogs

The Behaviour Analyst Certification Board (BACB) does not consider the well-being of the child through policy or guidelines

In fact, in the entire Behavior Analyst Certification Board’s professional code of ethics, a 24 page document, there is only one small section which refers to the behaviour-change program, and most of that involves how the plan is documented.

It is interesting to me that this document uses the word “client” to refer to the learner and the employer (parents or care home) interchangeably, which means that when it discussed “client consent,” it does not specify whether the actual learner has to consent. 

The way it is written, the learner could give no consent at all and the analyst could still be within guidelines as long as the employer consented, regardless of the learner’s age, intellectual ability, or communicative capacity.

Only three subsections in the Behavior Analyst Certification Board’s professional code of ethics even address the wellbeing of the learner:

  1. Subsection 4.08 requires that reinforcement procedures must be preferred over punishment procedures.
  2. Subsection 4.09 requires that they use the least restrictive procedures necessary.  This is meaningless, however, because it does not define what is considered “restrictive” or lay out a clear guideline on this.
  3. Subsection 4.10 warns against harmful reinforcers which could adversely affect the client’s health (presumably the learner, not the parent).

None of those three ensure that the procedures will be humane, for a few reasons. 

First, “reinforcement” and “punishment” do not mean “good” and “bad” in behaviourism. 

When I used to train service dogs, for example, some of the older trainers still used the old-fashioned, no-longer-recommended method of training a dog to retrieve: they dug their fingernails into the sensitive tip of the dog’s ear until the dog screamed, while holding out a dumbbell.  As soon as the dog’s mouth closed on the dumbbell, they stopped pinching the ear.

I refused to use this technique and now, if I were to do so, I would be barred from the Certification Council of Professional Dog Trainers for doing it.  But – and it is important to understand this – it is a “reinforcement technique.”

The dog is rewarded for grabbing the dumbbell by the trainer ending the pain.  That makes it reinforcement.

If the CCPDT used the BACB’s code of ethics, I could use that method without breaking their code of conduct. 

The BACB says nothing about inflicting pain.  There’s nothing in the BACB ethics code says you can’t use electric shock.  In fact, it doesn’t say anything at all about what type of “aversives” are acceptable.

4.08 – Behavior analysts ensure that aversive procedures are accompanied by an increased level of training, supervision, and oversight.  Behavior analysts must evaluate the effectiveness of aversive procedures in a timely manner and modify the behavior-change program if it is ineffective.  Behavior analysts always include a plan to discontinue the use of aversive procedures.

As long as the aversive procedure is effective and accompanied with training and supervision, under the ABA model you could hypothetically do anything. 

In fact, in a 24-page document detailing ethical codes of practice for working with human beings, including children and disabled adults, the word “abuse” is used zero times.

The word “humane” is used zero times.

The word “positive” is used zero times.

All this despite the fact that using behaviourism on a human being – especially a child – can easily venture into the realm of emotional and physical abuse.  For example, the Ontario Association of Children’s Aid Societies defines emotional abuse as:

excessive, aggressive, or unreasonable demands that place expectations on a child beyond his or her capacity.  Emotional abuse includes constantly criticizing, teasing, belittling, insulting, rejecting, ignoring, or isolating the child. 

And yet the BACB does not warn against or establish a criteria regarding the use of ignoring to punish a child, which is actually a commonly recommended strategy in ABA. 

The BACB does not warn against asking a child to sit still or work beyond the capacity expected and recommended for a child of their age.

The BACB, in other words, considers virtually anything acceptable so long as it is within the law and well documented.

I find that incomprehensible.

 The Rights of All Living Things

All sentient beings deserve to have their needs met, and their lives enjoyable.  All sentient beings should be able to be themselves and still be cared for.  All sentient beings should be treated with kindness and respect.

Modern dog training takes its codes of practice from the recommendations of many animal welfare bodies, and it prioritizes the needs and emotional well being of the dogs.

I don’t believe ABA focuses on the emotional needs of autistic children.  Nor do I believe that it incorporates recommendations from child psychologists on basic needs such as unconditional love, the need to play, or the right to say no.

The emotional needs of children are too often left entirely out of discussions about autism.  This should be shocking to anyone who understands children, behaviour, or how emotions and relationships impact us.

Nor does the field listen to autistic people about autistic emotions, which may be different from those of neurotypical children but are worth of respect nonetheless.

Dog trainers understand that dogs need to chew and bark and dig, but ABA therapists don’t understand that autistic children need to repeat words and sentences, flap their hands, and sit quietly rocking in a corner when things get too much.

ABA assumes that the key to happiness is changing their behaviour to be more in line with non-autistic children. 

It focuses on training children by holding their sources of happiness hostage and using them as blackmail to get the children to meet goals which are not necessarily in the best interest of their emotional health.

And like I said…

I wouldn’t treat a dog that way.

Carol Millman B.Sc., RAHT, CPDT-KA, CTDI is an autistic dog trainer with seven years of post-secondary education and a decade of practical experience.  She apprenticed as an instructor at Pacific Assistance Dogs Society in Burnaby, BC, and now trains dogs privately.  She specializes in training assistance dogs for people with a wide range of visible and invisible disabilities.