BCBAs Respond To The Dog Trainer Who Called Out ABA49 min read

Earlier this year, we pub­lished an article which shook the autism world — “Is Aba Really Dog Training For Children? A Professional Dog Trainer Weighs In.”

Contributor Carol Millman has a B.Sc in Psychology, a diploma in Animal Health Technology from the University of Guelph, and is a Certified Professional Dog Trainer as well as being cer­ti­fied AnimalKind by the SPCA.

In her April article, she dis­cussed ABA, par­tic­u­larly the way it is man­aged and reg­u­lated, and how it com­pares to the ethics and atti­tudes among pro­fes­sional dog training bodies who also employ con­di­tioning and behav­iour mod­i­fi­ca­tion in their work with ani­mals.

You can read the orig­inal here.

Below you will be able to read a letter, pub­lished in full at their spe­cific request, by Dr. Thomas Zane and Katie Gorycki.

Carol Millman has inserted her own replies within the doc­u­ment, also with per­mis­sion from the letter-writers.

picture of a puppy and a small boy looking at each other. Licensed through Can Stock Photo Inc. / oksun70

Response to Millman’s article:

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

Katie Gorycki and Thomas Zane

            Recently, the article “Is ABA Really “Dog Training for Children”?  A Professional Dog Trainer Weighs In” by Carol Millman was posted on The Aspergian, a site in which authors cat­alog expe­ri­ences, insights, knowl­edge, tal­ents, and cre­ative pur­suits of “autis­tics.” Millman’s article trig­gered numerous com­ments and responses, such as the article “ABA Is Not “Dog Training for Children.” –A Behavior Analyst Weighs In.” by Amanda Pakutz. Pakutz dis­cussed note­worthy topics in Applied Behavior Analysis (ABA), such as rad­ical behav­iorism, the seven dimen­sions of ABA, and the code of ethics.

Yes, I read and responded to Ms. Pakutz at the time. I found it odd that she often cited as proof the very same facts that I, too, had cited in my cri­tique.

For example, she says:

Millman writes, “I don’t believe ABA focuses on the emo­tional needs of autistic chil­dren.” Correct, it does not.

ABA does not employ inter­ven­tions based on any person’s per­ceived emo­tions. Practitioners of ABA focus on empir­i­cally based behav­ioral obser­va­tions in dif­ferent phys­ical envi­ron­ments.

Amanda Pakutz, ABA is not Dog Training For Children — A Behaviour Analyst Weighs In

She also links to the same com­pli­ance code that I linked to.

Then she says:

Millman con­torts the com­pli­ance code later with equiv­o­ca­tion fal­lacies. She gen­er­al­izes that because the code does not include “pos­i­tive” and “humane,” behav­iourism on a human being — espe­cially a child — can easily ven­ture into the realm of emo­tional and phys­ical abuse. The BCBA com­pli­ance code clearly states the fol­lowing:

Behavior ana­lysts have a respon­si­bility to operate in the best interest of clients. The term client as used here is broadly applic­able to whomever behavior ana­lysts pro­vide ser­vices, whether an indi­vidual person (ser­vice recip­ient), a parent or guardian of a ser­vice recip­ient, an orga­ni­za­tional rep­re­sen­ta­tive, a public or pri­vate orga­ni­za­tion, a firm, or a cor­po­ra­tion.

- Amanda Pakutz, ABA is Not Dog Training For Children — A Behaviour Analyst Weighs In

I can’t under­stand how saying that the com­pli­ance code does not dif­fer­en­tiate between a child and a cor­po­ra­tion proves that it pro­tects against emo­tional and phys­ical abuse.

If any­thing, this is one of the aspects of the code that I roundly crit­i­cized as impre­cise and full of con­ve­nient loop­holes. Quoting one sec­tion that pro­vides those loop­holes won’t con­vince me that the loop­holes do not exist.

It seems Ms Pakutz and I agree on the facts and she just dis­agrees with the con­clu­sion I drew from them.

We are behavior ana­lysts (like Ms. Pakutz), and we sup­port her posi­tion. We hope our response to Millman’s article will fur­ther clarify some mis­in­for­ma­tion and any con­fu­sion regarding ABA ser­vices. Additionally, we will explain how the two fields share sim­ilar values.

Oh, good.

A major topic in Millman’s article was on the well­being of indi­vid­uals and how ABA does not account for this. However, like Millman, our field of behavior analysis would agree that a pri­mary goal is to pro­tect the well­being of indi­vid­uals. 

Then I’m sure you’ll be seeking to improve the com­pli­ance code to better meet your goal.

Millman’s inter­pre­ta­tion of the Behavior Analyst Certification Board’s (BACB) code of ethics for the well­being of the indi­vidual needs some tweaking. For instance, Millman stated that the BACB does not con­tain any infor­ma­tion about the wel­fare of the therapy’s recip­i­ents and that “only three sub­sec­tions in the Behavior Analyst Certification Board’s pro­fes­sional code of ethics even address the well­being of the learner.”

That’s cor­rect.

Specifically, I said, “while it goes into great detail regarding what cer­ti­f­i­cants may do regarding 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.”

I con­trasted it with the ethics code of the Certification Council of Professional Dog Trainers, which lays out clear guide­lines about which methods should be tried first, and in exactly what order, starting with ensuring the dog’s phys­ical and emo­tional needs are met.

I claimed that the ABA ethics code does not pro­hibit abu­sive tech­niques, and does not require informed con­sent from the learner.

Please, tell me where I went wrong.

However, the behavior analyst’s pri­mary endeavor is to improve the lives of others. Specifically, the code was devel­oped to pro­tect the well­being of the learner, which includes sev­eral areas within the code, such as code 1.0 (respon­sible con­duct of behavior ana­lysts).

Items relating to client within this code include: 1.02(a,b) (bound­aries of com­pe­tence), 1.03 (main­taining com­pe­tence through pro­fes­sional devel­op­ment, 1.04(a‑e) (integrity), 1.05(a‑f) (pro­fes­sional and sci­en­tific rela­tion­ships, 1.06(a‑d) (mul­tiple rela­tion­ships and con­flicts of interest), and 1.07 (a‑d) (exploita­tive rela­tion­ships).

Great, let’s start there.

As you say, code 1.0 involves pro­fes­sional stan­dards, such as not doing things you aren’t trained to do, ensuring you get con­tin­uing edu­ca­tion, avoiding con­flicts of inter­ests, etc.

Under this sec­tion of the code, an ABA ther­a­pist could not dis­crim­i­nate based on age, sex, race etc. They cannot accept gifts from or have sex with their clients.

All very impor­tant.

But I don’t see how any of that ensures that the ther­a­pist doesn’t emo­tion­ally or phys­i­cally abuse the child.

For example, if the person is trained in admin­is­tering elec­trical shocks as an aver­sive, fol­lows accepted pro­to­cols, and gets con­tin­uing edu­ca­tion and avoids con­flicts of interest, they can con­tinue to apply elec­trical shocks to the learner so long as they don’t accept presents from them or molest them.

Furthermore, code 2.0 (behavior ana­lysts’ respon­si­bility to clients), is specif­i­cally devoted to the wel­fare of the client. Subsections 2.01 (accepting clients), 2.02 (respon­si­bility), 2.03(a,b) con­sul­ta­tion, 2.04(a‑d) (third-party involve­ment in ser­vices, 2.05(a‑e) (rights and pre­rog­a­tives of clients), 2.06(a‑e) (main­taining con­fi­den­tiality), 2.07(a,b) (main­taining records), 2.08 (dis­clo­sures), 2.09(a‑d) (treatment/intervention effi­cacy), 2.10(a,b) (doc­u­menting pro­fes­sional work and research), 2.11(a,b) (records and data), 2.12(a‑d) (con­tracts, fees, and finan­cial arrange­ments), 2.13 (accu­racy in billing reports), 2.14 (refer­rals and fees), and 2.15(a‑e) (inter­rupting or dis­con­tin­uing ser­vices) all focus on the learner.

First of all, I don’t see how Section 2.0 is ded­i­cated to the well­being of the learner con­sid­ering that this sec­tion opens by reminding us that “client” can mean the employer, not nec­es­sarily the learner.

Under the sec­tion on respon­si­bility, the code says that if mul­tiple people could be con­sid­ered the client, the behavior ana­lyst gets to pick who they think the “ulti­mate ben­e­fi­ciary of ser­vices is.”

Now, if the child melts down in gro­cery stores and the goal is to get the child to be quiet so mama can do her shop­ping, I would argue that the ben­e­fi­ciary is not the child, who still has to go to the loud, busy, over­whelming gro­cery store, but the mother, whose life has been made so much easier.

In my work, the ben­e­fi­ciary of my training is very often the owner, not the dog. The dog is quite happy stealing food off of tables and run­ning amok on the trails. I could argue that I’m saving the dog from being rehomed, but I’m sure the dog wouldn’t see it that way.

It’s not the dog’s fault that it’s acting like a dog and its owners don’t know how to handle that. I do my job for my clients, who are the owners, so they can enjoy their dog’s pres­ence in their lives.

Nothing in this sec­tion addresses the use of aver­sives, the use of neg­a­tive rein­force­ment, pos­i­tive pun­ish­ment extinc­tion, or any­thing else in that vein. They are basic pro­fes­sional stan­dards and could just as easily be guide­lines for lawyers or bankers.

In fact, it could be right out of a tem­plate, hon­estly.

My com­plaint was very spe­cific. I didn’t call behav­iour ana­lysts ter­rible mon­sters who set out to destroy chil­dren’s lives. I’m sure they do want to improve lives. I just don’t think that they focus their goals on the child’s internal emo­tional expe­ri­ence.

I don’t think that’s an out­ra­geous claim con­sid­ering that modern ABA texts still use the “black box,” anti-mentalism stance of 1960s behav­iourism.

I accused ABA of using out­dated forms of behav­iourism, of pri­or­i­tizing “skills” over the sub­jec­tive internal expe­ri­ence of its recip­i­ents, and I said that many of its prac­tices and poli­cies risk crossing the bound­aries of emo­tional and some­times even phys­ical abuse.

I said, “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.”

I don’t believe that ABA rec­og­nizes the limits of behav­iourism or teaches its ther­a­pists and ana­lysts about these limits or about more modern dis­cov­eries in psy­chology and neu­ro­science. Most behavior ana­lysts ignore sen­sory pro­cessing dis­order, for example, with many directly calling it pseu­do­science.

You can’t con­vince me oth­er­wise by listing guide­lines about pro­fes­sional con­duct regarding the person who hired you.

Section 3.0 (assessing behavior) con­tinues to includes spe­cific oblig­a­tions behavior ana­lysts must meet that focus on behavior ana­lytic tech­niques that ben­efit the learner. Items relating to this code are: 3.01(a,b) (behavior-analytics assess­ment), 3.02 (med­ical con­sul­ta­tion), 3.03 (a,b) (behavior-analytic assess­ment con­sent), 3.04 (explaining assess­ment results, and 3.05 (consent-client records).

Okay, now we’re get­ting out of the basic stuff and into more ABA-specific guide­lines.

The sec­tion on Analytic Assessment, 3.01, is shock­ingly short.

Unlike the Humane Hierarchy and LIMA stan­dards laid out by the CCPDT, this ethics code leaves quite a lot of decision-making in the ana­lyst’s hands.

The type of assess­ment used is deter­mined by client’s needs and con­sent, envi­ron­mental para­me­ters, and other con­tex­tual vari­ables.

The code does not lay out what type of assess­ments should be used based on which needs and vari­ables. It’s up to the ana­lyst to decide, and what­ever they decide could be jus­ti­fied under this guide­line.

But it does specify that they need to graph the data… so you know, that’s some­thing.

Regarding con­sent, it simply requires “the client’s” written con­sent. It doesn’t specify that the learner must con­sent. Remember that “the client” could mean the parent or a freaking cor­po­ra­tion.

Now, since many of their learners are non­verbal and have not been taught or given access to AAC, or may be minors, I would be fine with that if the code made more spe­cific dis­tinc­tions between the learner and the parent, and guide­lines regarding the age of con­sent.

Perhaps they could look toward med­ical pae­di­atrics for guide­lines.

But as it is written, a child or an adult could have goals set for them, rewards and pun­ish­ments inflicted on them, all without their con­sent.

Code 4.0 (behavior ana­lysts and the behavior-change pro­gram) also include mul­tiple areas relating to the well­being of the learner. This includes 4.02 (involving clients in plan­ning and con­sent), 4.03(a,b) (indi­vid­u­al­ized behavior-change pro­grams), 4.04 (approving behavior-change pro­grams), 4.05 (describing behavior-change pro­gram objec­tives, 4.06 (describing con­di­tions for behavior-change pro­gram suc­cess), 4.07(a,b) (envi­ron­mental con­di­tions that inter­fere with imple­men­ta­tion), 4.08(a‑d) (con­sid­er­a­tions regarding pun­ish­ment pro­ce­dures), 4.09 (least restric­tive pro­ce­dures), 4.10 (avoiding harmful rein­forcers), and 4.11(a,b) (dis­con­tin­uing behavior-change pro­grams and behavior-analytic ser­vices).

I’m get­ting a little tired of these lists.

Like, I’ve read the code over sev­eral times. Listing the titles of the sub­sec­tions that I have already read doesn’t seem per­sua­sive or useful.

You know what?

Let’s save everyone a click and just copy and paste this whole sec­tion for our readers to judge for them­selves.

4.01 Conceptual Consistency.

Behavior ana­lysts design behavior-change pro­grams that are con­cep­tu­ally con­sis­tent with behav­ior­an­a­lytic prin­ci­ples.


Does this rule pro­hibit electro-shocking a child, leaving them in iso­la­tion, ignoring them during a melt­down, forcing them to work for 40 hours a week, or stop­ping them from per­forming a soothing stim?


4.02 Involving Clients in Planning and Consent.

Behavior ana­lysts involve the client in the plan­ning of and con­sent for behavior-change pro­grams.

What do you think? Does THAT rule pro­hibit electro-shocking a child, leaving them in iso­la­tion, ignoring them during a melt­down, forcing them to work for 40 hours a week, or stop­ping them from per­forming a soothing stim?

I don’t see how, con­sid­ering that “the client” can mean a parent who has been told (or even court-ordered) that this step is nec­es­sary to help their child.

4.03 Individualized Behavior-Change Programs.

(a) Behavior ana­lysts must tailor behavior-change pro­grams to the unique behav­iors, envi­ron­mental vari­ables, assess­ment results, and goals of each client. (b) Behavior ana­lysts do not pla­gia­rize other pro­fes­sionals’ behavior-change pro­grams.

Ah, ok, well in this case it could be wrong to electro-shock a child, leave them in iso­la­tion, ignore them during a melt­down, force them to work for 40 hours a week, or stop them from per­forming a soothing stim if it is a pla­gia­rized behav­iour change pro­gram from another pro­fes­sional.

4.04 Approving Behavior-Change Programs.

Behavior ana­lysts must obtain the client’s written approval of the behavior-change pro­gram before imple­men­ta­tion or making sig­nif­i­cant mod­i­fi­ca­tions (e.g., change in goals, use of new pro­ce­dures).

I see, so as long as the cor­po­ra­tion employing you signs off, it’s okay.

4.05 Describing Behavior-Change Program Objectives.

Behavior ana­lysts describe, in writing, the objec­tives of the behavior-change pro­gram to the client before attempting to imple­ment the pro­gram. To the extent pos­sible, a risk-benefit analysis should be con­ducted on the pro­ce­dures to be imple­mented to reach the objec­tive. The descrip­tion of pro­gram objec­tives and the means by which they will be accom­plished is an ongoing process throughout the dura­tion of the client-practitioner rela­tion­ship.

This rule seems to lay out exactly how you could go about electro-shocking a child, leaving them in iso­la­tion, ignoring them, forcing them to work for 40 hours a week, or stop­ping them from per­forming a soothing stim.

You just need to have a risk-benefit analysis sheet which deter­mines that these things are nec­es­sary to achieve the goals of the pro­gram. If the client is a cor­po­ra­tion then the ben­efit of 40 hours per week might be jus­ti­fi­able as com­pared to the risk of trauma to the devel­oping child.

4.06 Describing Conditions for Behavior-Change Program Success.

Behavior ana­lysts describe to the client the envi­ron­mental con­di­tions that are nec­es­sary for the behavior-change pro­gram to be effec­tive.

Okay. So as long as they say “we need an iso­la­tion room for planned ignoring” you can ignore a child.

4.07 Environmental Conditions that Interfere with Implementation.

(a) If envi­ron­mental con­di­tions pre­vent imple­men­ta­tion of a behavior-change pro­gram, behavior ana­lysts rec­om­mend that other pro­fes­sional assis­tance (e.g., assess­ment, con­sul­ta­tion or ther­a­peutic inter­ven­tion by other pro­fes­sionals) be sought. (b) If envi­ron­mental con­di­tions hinder imple­men­ta­tion of the behavior-change pro­gram, behavior ana­lysts seek to elim­i­nate the envi­ron­mental con­straints, or iden­tify in writing the obsta­cles to doing so.

My brain is doing funny things and imag­ining the ana­lysts writing sternly worded let­ters to the weather about envi­ron­mental con­di­tions.

(No, I don’t think that’s what they mean. Yes, I know what they mean. No, I don’t think it pro­hibits poten­tial abuse.)

4.08 Considerations Regarding Punishment Procedures.

(a) Behavior ana­lysts rec­om­mend rein­force­ment rather than pun­ish­ment when­ever pos­sible. (b) If pun­ish­ment pro­ce­dures are nec­es­sary, behavior ana­lysts always include rein­force­ment pro­ce­dures for alter­na­tive behavior in the behavior-change pro­gram. © Before imple­menting punishment-based pro­ce­dures, behavior ana­lysts ensure that appro­priate steps have been taken to imple­ment reinforcement-based pro­ce­dures unless the severity or dan­ger­ous­ness of the behavior neces­si­tates imme­diate use of aver­sive pro­ce­dures. (d) 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 when no longer needed.

There it is, in plain writing — pun­ish­ment is per­mitted, as are aver­sives, as long as you have taken “appro­priate steps” to imple­ment reinforcement-based pro­ce­dures and oversee the pro­ceed­ings to make sure it’s working.

It would be nice to know what “appro­priate steps” are.

Positive rein­force­ment first? How many times? How many dif­ferent rein­forcers must be tried? What steps should be taken to ensure that this behav­iour is even oper­antly medi­ated and not a reflex/something out­side the learn­er’s con­trol?

What about the goal involved? What goals are worth pun­ish­ment pro­ceed­ings? Which goals are not? Or are all goals cre­ated equal?

What kinds of pun­ish­ments? What kinds of aver­sives? Why is there carte blanche?

But the bigger ques­tion is, why are aver­sives per­mitted in any cir­cum­stances?

It says “if they are nec­es­sary” but it doesn’t define what sit­u­a­tions could count as “nec­es­sary,” leaving it open to inter­pre­ta­tion.

Reams of studies on ani­mals (from rats to dogs to birds), as well as moun­tains of research on chil­dren have proven that aver­sives and pun­ish­ments cause trauma, increase behav­iour prob­lems, raise cor­tisol levels, and are inef­fec­tive in the long term.

Studies on dogs have revealed that dogs trained using pun­ish­ment are more likely to develop aggres­sive behav­iours.

In dog training, aver­sives include choke chains, pinch col­lars, unpleasant noises, unpleasant smells, spray bot­tles, vinegar, and elec­tric shocks.

What kind aver­sives does one use on a child?

Could you spritz them with vinegar and water if they don’t say “please”? I mean, assuming they tried rein­force­ment first and then decided the aver­sive was “nec­es­sary”?

What about neg­a­tive pun­ish­ment — can you deny the child food? Love? Free move­ment? Communication boards?

What can a child be denied, and for what types of goals?

Why doesn’t this sec­tion lay out what types of pun­ish­ments can be used and which cannot?

B.F. Skinner was scornful about the use of pun­ish­ment. Any good behav­iourist knows that pun­ish­ment is use­less in the long term.

Punishment cre­ates avoid­ance of the pun­isher, but the organism will con­tinue to attempt to return to the behav­iour and once the pun­ish­ment is dis­con­tinued, the behav­iour re-emerges.

You must know this.

So why is it even per­mitted on human beings?

Corporal pun­ish­ment, denial of basic rights like food and access to the care­giver, and even just authoriarian-style par­enting are con­nected heavily to neg­a­tive out­comes in the child’s mental health.

Long term, it has no effect.

So why aren’t those things pro­hib­ited out­right — explic­itly and in great detail?

Why does this sec­tion leave the ana­lyst free to apply aver­sives of any type to human beings? Free to deny basic rights like food or com­mu­ni­ca­tion?

My SPCA AnimalKind cer­ti­fi­ca­tion pro­hibits me out­right from using aver­sives.

Their cer­ti­fi­ca­tion stan­dards are even more strin­gent than the CCPDT, which I used to con­trast against your com­pli­ance code. They included observing me during sev­eral training ses­sions, pouring through my records and stan­dard oper­ating pro­ce­dures, and asking to see our diplomas.

I used the CCPDT’s guide­lines for com­par­ison because I didn’t think it would be fair to hold ABA to the same stan­dards as the SPCA.

After all, everyone knows that people are much more con­cerned about animal wel­fare than human wel­fare.

My point is that aver­sives are cruel and inef­fec­tive. I think it’s wrong that they are even used on ani­mals let alone human chil­dren.

How dare you list this sec­tion as proof that the learn­er’s well-being is in any way pro­tected?

4.09 Least Restrictive Procedures.

Behavior ana­lysts review and appraise the restric­tive­ness of pro­ce­dures and always rec­om­mend the least restric­tive pro­ce­dures likely to be effec­tive.

So there’s no limit on how restric­tive a pro­ce­dure can be, so long as it is “likely” nec­ces­sary to be “effec­tive,” eh?

The end jus­ti­fies the means, clearly.

4.10 Avoiding Harmful Reinforcers.

Behavior ana­lysts min­i­mize the use of items as poten­tial rein­forcers that may be harmful to the health and devel­op­ment of the client, or that may require exces­sive moti­vating oper­a­tions to be effec­tive.

What does this even mean?

Don’t give the kid candy?

You have a sec­tion on avoiding harmful REWARDS but nothing on avoiding harmful pun­ish­ments? What if the aver­sives are harmful to the health and devel­op­ment of the client? That’s fine, then?

Or are you dis­cussing neg­a­tive rein­forcers, such as the end of pain?

In that case the rein­force­ment is the end of the pain, so of course that isn’t harmful but the pain sure as hell would be. That’s what pain is — a sign of harm.

So… say you agree to dim the lights (which are hurting the learn­er’s eyes and trig­gering a melt­down) only once the learner com­plies to your demand.

Sorry. “Mand.”

Dimming the lights of course is not a harmful rein­forcer so this pro­tocol would be just fine, then?

In this case, isn’t refusing to dim the lights causing harm, because you are extending their dis­com­fort?

Where is that cov­ered in the com­pli­ance code?

And what if the rein­forcer is some­thing the child has a basic right to? Like their AAC device, or their comfort-object, or their meals? Is it okay to use those as rein­forcers?

Because it shouldn’t be.

I never make a dog work for water. I never starve a dog to get com­pli­ance. I never force a dog to stay in a con­fined space if the dog is dis­tressed. That’s basic humane treat­ment and respect for the learner as an active par­tic­i­pant and not a slave.

But I see nothing in these guide­lines to stop me from doing these things to a child.

Why is it per­miss­able within the code to make access to your AAC device con­tin­gent on behav­iour? To deprive a child of their voice because they weren’t com­pliant enough?

This hap­pens reg­u­larly — many non­speakers tell of their iPad being taken away for “mis­be­hav­iour” or “non­com­pli­ance,” and example videos on YouTube demon­strate it being done in between dis­crete trials.

This video, which was fea­tured in C.L. Lynch’s article on ABA and abuse, shows the AAC device being repeat­edly removed from the child’s hands, thus depriving him of a voice except during very brief inter­ludes.

Why is this per­mitted?

Why aren’t there guide­lines on this?

But returning to your letter…

   In addi­tion,  there are sev­eral over­ar­ching core eth­ical prin­ci­ples within behavior analysis, addressing uni­versal human right con­cerns, including doing no harm, respecting autonomy, ben­e­fiting others, being just, being faithful, according to dig­nity, treating others with caring and com­pas­sion, the pur­suit of excel­lence, and accepting account­ability. (Bailey & Burch, 2016). It is a require­ment for behavior ana­lysts to under­stand these prin­ci­ples within the code and is essen­tial in order to pro­vide respect to the field of ABA ser­vices. Additionally, six rights out­lined by Van Houten et al. 1998, specify that an indi­vidual has a right to (1) a ther­a­peutic envi­ron­ment, (2) ser­vices whose over­riding goal is per­sonal wel­fare, (3) treat­ment by a com­pe­tent behavior ana­lyst, (4) pro­grams that teach func­tional skills, (5) behav­ioral assess­ment and ongoing eval­u­a­tion, and (6) the most effec­tive treat­ment pro­ce­dures avail­able. The impor­tance of these rights described by Van Houten et al. 1988 is that “the field of behavior analysis acknowl­edges its respon­si­bil­i­ties by reaf­firming its con­cern for indi­vidual wel­fare and by pre­scribing the means by which behav­ioral treat­ment can be deliv­ered in the most ben­e­fi­cial manner (pp. 384).” Additionally, Van Houten et al. assessed both the eth­ical and appro­priate appli­ca­tion of behavior ana­lytic treat­ment of clients’ rights.


Then please explain to me how it is pos­sible that the BCBAs who starve and elec­tro­cute people at the Judge Rotenberg Center are not only still mem­bers in good standing but are even invited to speak at ABAI con­fer­ences.

In my article, The Shocking Truth About ABA, Autism and Abuse, I quoted an inmate of that center.

“I got shocked for tic like body move­ments, for which I have no con­trol over, and which don’t hurt me or any­body else. I would be shocked for waving my hand in front of my face for more than 5 sec­onds, for closing my ears with my fin­gers, which I do when things get too loud, because I cannot tol­erate too much noise.”

Judge Rotenberg Survivor Letter

If your ethics code ensures a ther­a­peutic envi­ron­ment where the rights of the learner are pri­or­i­tized, how is it pos­sible that the people who did this are not only mem­bers in good standing, but being invited to speak at ABA con­fer­ences?

To speak on ethics of all things?

Why isn’t there an outcry among other BCBAs? Why aren’t you fighting to have this kind of thing con­demned?

Why are you, instead, arguing with a dog trainer on the internet about whether or not the existing Code is accept­able?

Can’t you see that it is not, if chil­dren can be shocked for cov­ering their ears without the ana­lyst losing their cre­den­tials?

The fact is, ABA believes that the end jus­ti­fies the means. If you can point to a graph that shows inci­dences of a behav­iour going down, you can jus­tify elec­tro­cuting a help­less human being under your care.

But that doesn’t mean the person feels they have been helped.

It doesn’t ensure that if the learners beg you to stop, that you will stop.

Any number of atroc­i­ties could be per­formed in the name of ABA so long as it is doc­u­mented, graphed, and approved by someone — not nec­es­sarily the learner.

When dis­cussing the his­tory of applied behavior analysis, we under­stand Millman’s con­cerns regarding the neg­a­tive aspects of treat­ments used at the begin­ning of our field in the mid-1960s. However, it must be noted that although the his­tory of our field included treat­ments that are not viewed as accept­able today, we cre­ated a code of Ethics to pre­vent behavior ana­lysts’ imple­men­ta­tion of uneth­ical treat­ments. As Bailey and Burch (2016) noted, the evo­lu­tion of our code of Ethics has put in place all the nec­es­sary ele­ments of con­trol and man­age­ment to min­i­mize future abuses. It is impor­tant to under­stand that the field of ABA has moved for­ward from those ear­lier times with ques­tion­able ethics.

The Judge Rotenberg Center is still in oper­a­tion.

This is not old news. This is cur­rent events. They are starving people, elec­tro­cuting people, scar­ring people, and they are being invited to con­fer­ences.

Don’t tell me the days of Lovaas are over when mem­bers in good standing are still pulling the same old tricks.

Don’t tell me that your code of Ethics has put in place all the nec­es­sary ele­ments to min­i­mize abuses when your same code of ethics allows this:

“My meals came to the class­room cut into tiny pieces and divided into por­tions inside of a little plastic cup. Every time I had one of these little behav­iors, it forced me to stand up and throw one cup away. There were many days I would lose most of my meals. And the hun­grier I got, the more frantic and rest­less my body became. This caused me to have more behav­iors like tics and rocking, and in turn I would lose more food.” 

Judge Rotenberg Center Survivor Letter

Despite some neg­a­tive aspects of the his­tory of our field, one cannot ignore the pos­i­tives that have come from the past. For instance, one of our founding fathers, B.F. Skinner, won the Humanist Award in 1972. Then, Skinner wrote “…If Humanism meant nothing more than the max­i­mizing of per­sonal freedom and dig­nity, then I was not a Humanist. If it meant trying to save the human species, then I was (Skinner, 1983, pp. 343).” Additionally, Skinner’s pri­mary con­cern was inde­pen­dence of the indi­vidual, from the usual neg­a­tive con­tin­gen­cies incor­po­rated by the culture’s insti­tu­tions and cus­toms.  We con­tinue to share that goal today.

I’m a fan of Skinner, per­son­ally, and I think he’d think your code of ethics is garbage. Sorry.

He was against pun­ish­ment because he knew how utterly use­less it is.

Severe pun­ish­ment unques­tion­ably has an imme­diate effect in reducing a ten­dency to act in a given way. This result is no doubt respon­sible for its wide­spread use. We ‘instinc­tively’ attack anyone whose behavior dis­pleases us — per­haps not in phys­ical assault, but with crit­i­cism, dis­ap­proval, blame, or ridicule.

Whether or not there is an inher­ited ten­dency to do this, the imme­diate effect of the prac­tice is rein­forcing enough to explain its cur­rency. In the long run, how­ever, pun­ish­ment does not actu­ally elim­i­nate behavior from a reper­toire, and its tem­po­rary achieve­ment is obtained at tremen­dous cost in reducing the over-all effi­ciency and hap­pi­ness of the group.

B.F. Skinner

Given Skinner’s stance on pun­ish­ment, once again I want to express amaze­ment that your code of ethics even per­mits its use in any con­text what­so­ever.

The treat­ment of humans through ABA ser­vices was addressed in a com­ment by Millman to Pakutz’s article, which stated “If you truly care about autistic people, then instead of defending a very loose code of con­duct which allows ther­a­pists to do things like this: TW: Judge Rotenberg Center (JRC), elec­tric shock, abuse…As you can see, sev­eral of the board mem­bers at the JRC are cer­ti­fied by the BCBA (Millman, 2019).” It is impor­tant to note that JRC does not rep­re­sent our entire field, just as cases of animal abuse in your field does not dis­credit the entire field because of a few cases of animal abuse.

Oh no no no.

No no no no no no.

You can’t point to the JRC and com­pare it to “a few cases of animal abuse.”

We aren’t talking about some rogue people claiming to be ana­lysts just because they watched The Autist Whisperer on NatGeoWild.

The ana­lysts at the JRC are edu­cated in ABA and are cer­ti­f­i­cants in good standing.

They are being invited to speak at con­fer­ences.

I guar­antee you that if the CCPDT, or the BCSPCA ‑both bodies who have cer­ti­fied me and my methods- found out that I was rou­tinely elec­tro­cuting dogs or starving them into good behav­iour, I would not be a member in good standing.

I would be out on my ass.

They would pub­lish my name on their web­site along with the reason for my decer­ti­fi­ca­tion.

I would 100% not be per­mitted to speak at one of their con­fer­ences.

Additionally, in response to Millman’s com­ment on JRC’s use of shock therapy, it is note­worthy to point out that the majority of our field does not accept the use of pun­ish­ment and it is only used as a last resort pro­ce­dure (Bailey & Burch, 2016). We under­stand Millman’s con­cern that mem­bers from JRC pre­sented at the Association for Behavior Analysis International’s (ABAI) con­ven­tion. Although this is true, it should be empha­sized that a panel in our field protested JRC at our most recent con­fer­ence in Chicago last month, to renounce shock tor­ture as part of treat­ment.

Okay, so the entire pop­u­la­tion of BCBAs don’t like the idea of shocking people.

That’s good. The ana­lysts who protested get a click and a treat.

Were you one of them?

But it doesn’t change the fact that it is per­mitted.

Either your code of con­duct per­mits these abuses… or it does not.

Either your code of con­duct is suf­fi­cient… or it is not.

The dog trainers who shock their dogs, use out­dated “dom­i­nance” non­sense, or end up in court because they heli­coptered a dog and broke its spine, are not cer­ti­f­i­cants of the CCPDT.

In fact, even though the CCPDT has made it vir­tu­ally impos­sible to use aver­sives on a dog through their guide­lines and have pub­lished a posi­tion state­ment on shock as an aver­sive threat­ening dire con­se­quences, many people are unhappy that it is even hypo­thet­i­cally per­mis­sible and I pre­dict this will change in the near future.

If you were against shock, you’d be saying “you’re right, it shouldn’t be per­mis­sible, this com­pli­ance code needs improve­ment.”

Worse — the fact that the code per­mits it and that the BCBAs who prac­tise it are still mem­bers in good standing is directly respon­sible for the con­tinued exis­tence of that facility.

That makes them not just accepting but com­plicit.

The judge who heard the case against it decided in favour of the JRC because it was made clear that the elec­tric shocks “con­form to the accepted stan­dard of care for treating indi­vid­uals with intel­lec­tual and devel­op­mental dis­abil­i­ties.”

Speaking as someone with a diag­nosed neu­rode­vel­op­mental dis­ability, I’m extremely unhappy about this.

All of those sec­tions you quoted at me are totally use­less if they cannot pro­tect someone from being shocked or starved because they made a noise or fid­geted too much.

All of those sec­tions are totally use­less if they result in “level III aver­sives” being upheld in state court because they “con­form” to “accepted stan­dards of care.”

I’m sorry that I’m holding your asso­ci­a­tion to high stan­dards — like the stan­dards used by mul­tiple cer­ti­fying bodies for people who work with ani­mals — but the cer­ti­fying bodies have a duty of care to the learners, and if they are accepting pro­ce­dures which are described as tor­ture by the United Nations, then they are not doing their job.

If it isn’t okay to shock a trai­torous spy into com­pli­ance, it isn’t okay to do it to someone with an intel­lec­tual dis­ability.

You can’t say that the asso­ci­a­tion is pro­tecting chil­dren because MOST ana­lysts wouldn’t abuse chil­dren. It shouldn’t be left up to the indi­vidual ana­lyst’s con­science.

That’s anarchy, not reg­u­la­tion.

Either your field is closely reg­u­lated, tightly con­trolled, and based on up-to-date sci­ence sur­rounding psy­chology or it isn’t, and ana­lysts are free to do what­ever they like so long as they mind their p’s and q’s and can pro­duce jus­ti­fi­ca­tions for their actions.

…So which is it?

And yes, I point to the JRC as an extreme example of the kinds of abuses which are per­mitted under your Compliance Code, but it needn’t be any­thing so extreme.

Mundane everyday abuse occurs in the name of ABA all over the world every day, because indi­vidual ana­lysts are given the power to make their own deci­sions regarding neg­a­tive rein­force­ment, pos­i­tive pun­ish­ment, and goal-setting without any strict con­trols or defin­i­tive guide­lines.

Questions like, “Is it okay to force a child to remain in the painful sit­u­a­tion (be it gro­cery store, loud envi­ron­ment, hands being ‚what­ever it might be) until they comply to avoid rein­forcing neg­a­tive behavior?” should not be left up to the indi­vidual ana­lyst.

That leaves them open to deciding that the answer is yes, in which case they are enforcing pain on a child for the sake of teaching them to shut up and be quiet.

That’s not pri­or­i­tizing the child’s wel­fare. It is causing harm.

If an ana­lyst can jus­tify causing harm in the name of neg­a­tive rein­force­ment, then the com­pli­ance code pri­or­i­tizes arbi­trary goals over well­being of the learner.

It’s that simple.

If a dog is becoming dis­tressed and howls in his crate for hours on end, I rec­om­mend let­ting the dog out.

Is that rein­forcing the howling? Sure. But the dog is clearly dis­tressed, and that comes first.

If I put this dog into a sit­u­a­tion he wasn’t ready to handle, that’s on ME. He isn’t howling because he’s a bad dog. He’s howling because I messed up.

I am a skilled enough trainer that I can get that dog back in the crate and being quiet without risking building a trauma his­tory.

And if he really hates it that much? We’ll find a dif­ferent solu­tion to our problem — one that doesn’t require a crate.

That deci­sion isn’t left up to me and my con­science, either.

Allowing a dog to howl and cry in its crate until it stops, come what may, would be neg­a­tive reinforcement/extinction (depending on the pro­tocol) which is marked by a big yellow cau­tion sign.

I can’t do that until I have first been through four other steps, by which time I would def­i­nitely have that dog com­fort­able in the crate.

I’m not so lim­ited in skill and imag­i­na­tion that I could ever jus­tify con­fining a crying, howling dog at all costs. I’ll let a tired, sleepy puppy whine for a few min­utes before falling asleep. That’s it.

And yet how many autistic people remember being locked in a room to cry and howl for hours?

Too many.

And they were the lucky ones, because at least they didn’t have aver­sives used on them.

Questions like, “Is it appro­priate to remove the iPad he uses to com­mu­ni­cate as a result of mis­be­hav­iour?” should also not be left up to the indi­vidual ana­lyst.

You are removing a child’s voice. That is denying basic human rights.

I never deny a dog their basic rights.

They always have water. They are always wel­come to leave the sit­u­a­tion if they need to.

My learners par­tic­i­pate vol­un­tarily, and if they have some­thing to say, then I don’t punish them. Sometimes they get excited or frus­trated and need to yap. Or they get scared and need to snarl and growl. I let them have their say and then ask if they want to go back to work.

If they don’t, then we dis­con­tinue and just share treats and cud­dles. Does that reward the avoidant behav­iour? Maybe. But that’s not my biggest pri­ority.

In addi­tion, we take excep­tion with Millman’s mis­con­cep­tion of ABA’s use of pun­ish­ment in the field, since most of our field does not accept the use of pun­ish­ment. Millman’s state­ment that “the use of ignoring to punish a child, which is actu­ally a com­monly rec­om­mended strategy in ABA,” is not accu­rate. Subsection 4.08 (con­sid­er­a­tions regarding pun­ish­ment pro­ce­dures) in the BACB pro­fes­sional and eth­ical com­pli­ance code spec­i­fies that (a) behavior ana­lysts rec­om­mend rein­force­ment rather than pun­ish­ment when­ever pos­sible, (b) if pun­ish­ment pro­ce­dures are nec­es­sary, behavior ana­lysts always include rein­force­ment pro­ce­dures for alter­na­tive behavior in the behavior-change pro­gram, © before imple­menting punishment-based pro­ce­dures, behavior ana­lysts ensure that appro­priate steps have been taken to imple­ment reinforcement-based pro­ce­dures unless the severity or dan­ger­ous­ness of the behavior neces­si­tates imme­diate use of aver­sive pro­ce­dures, and (d) 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 when no longer needed. Behavior ana­lysts always include a plan to dis­con­tinue the use of aver­sive pro­ce­dures when no longer needed(BACB, 2019, pp. 12–13)”.

Okay, well, first of all, your code men­tions nothing about extinc­tion. Second of all, telling me what the code says doesn’t prove that this prac­tice is, in fact, uncommon.

Are you really claiming that ignoring is not a common tactic in ABA?

Do you want all of the autistic people who were sub­jected to it or have wit­nessed it to tell you about their expe­ri­ences in the com­ment sec­tion?

If I poll ABA ther­a­pists in my area, will they all tell me that ignoring is never or very rarely done?

…Or will they tell me some­thing else?

Because I’ve got to tell you, I’ve read ABA texts and they describe the use of extinc­tion through ignoring pretty reg­u­larly, and there were no pretty graphics showing yellow cau­tion signs either.

It is well estab­lished that pun­ish­ment may pro­duce trou­ble­some side effects (Cooper, Heron, & Heward, 2007, pp. 336–338), and other strate­gies should almost always be used first. However, there may be some life-threatening sit­u­a­tions in which pun­ish­ment as a first option would be nec­es­sary, but these are extremely rare. It should be clear that if pun­ish­ment pro­ce­dures are nec­es­sary, the behavior ana­lyst must first con­duct a func­tional assess­ment, to iden­tify con­trol­ling vari­ables, and if alter­na­tive behav­iors are iden­ti­fied, then rein­forcers for these behav­iors prob­ably can be found as well. Finally, an eth­ical behavior ana­lyst will always con­sider the pos­si­bility of med­ical or bio­log­ical fac­tors that may elicit dan­gerous behav­iors. In sum, we strive to avoid using pun­ish­ment at all costs.

We have a saying in dog training — “More skill means less force.” The need to apply force and pos­i­tive pun­ish­ment (aver­sives) is a flag that the trainer is unskilled.

A skilled trainer never needs to use force, because they know how to cor­rectly diag­nose an issue and cor­rect it.

Recently, a dog trainer near me made the news for repeat­edly punching a dog in the head. Do you know what his own vet said?

“Punching and holding a dog to the ground is some­times nec­es­sary.”

This is an animal pro­fes­sional who thinks that punching a dog in the head is “some­times nec­es­sary.” What that really means is that he doesn’t know any other way.

I do.

I know a lot of other ways.

That’s why I’m a cer­ti­fied dog trainer, and he is not.

The number of other, more sci­en­tif­i­cally valid as well as humane, ways could — and do — fill sev­eral books.

But you shouldn’t be leaving some­thing like this up to the indi­vidual skill of the ana­lyst. Once again, the com­pli­ance code should lay out clearly exactly which steps must be tried first and in what order. It must define what “nec­es­sary” means.

What goals are worth pun­ish­ment to achieve? What goals are not best achieved through more effec­tive and more humane tac­tics?

Very few, if any, in my mind.

In regard to sub­sec­tion 4.09, Millman stated that we “use the least restric­tive pro­ce­dures nec­es­sary.  This is mean­ing­less, how­ever, because it does not define what is con­sid­ered “restric­tive” or layout a clear guide­line on this.” We would like to pro­vide a def­i­n­i­tion to pro­vide a clearer guide­line on the use of restric­tive pro­ce­dures. Bailey and Burch (2016) define a restric­tive pro­ce­dure as “a prac­tice that limits an indi­vid­u­al’s move­ment, activity or func­tion; inter­feres with an indi­vid­u­al’s ability to acquire pos­i­tive rein­force­ment; results in the loss of objects or activ­i­ties that an indi­vidual values; or requires an indi­vidual to engage in given freedom of choice (Bailey and Burch, 2016, pp.135).”  This code directs behavior ana­lysts to first use the least restric­tive methods that do not inter­fere with the client’s ability to con­tact rein­forcers and pro­vides the client freedom of choice.

But my point is that it doesn’t pro­vide clear guide­lines.

Where is the line drawn in the sand? Is it okay to phys­i­cally restrain a child? Using which types of holds? How about tying the child? Straight jackets? Grabbing the hands and folding them into their lap?

Is it okay to tie them? Lock them up? Why are there no hard-and-fast guide­lines regarding what is accept­able and what is not?

I don’t want to hear that it is better not to tie a child up. I want to hear that it is for­bidden to tie a child up.

A cer­tain amount of Googling with just the right key­words helped me dig up a posi­tion stat­ment from ABA International on restraint.

At first blush it looks okay.

ABAI is opposed to the use of seclu­sion when it is oper­a­tionally defined as placing someone in a locked room, often com­bined with the use of mechan­ical restraint and/or seda­tion.…

ABA International Positions and Policies

That sounds good…

…and not part of a formal Behavior Intervention Plan to which the indi­vidual served and/or their Guardians have con­sented. 

ABA International Positions and Policies


Seclusion is oper­a­tionally defined only as some­thing that is NOT part of a formal Behaviour Intervention Plan? So… if it IS, and the guardian con­sented… then it isn’t seclu­sion? And the guardian can be a foster parent, or social ser­vices, and the child can be a victim of trauma and abuse?


The use of restraint in a planned Behavior Intervention Plan is done as part of an inte­grated effort to reduce the future prob­a­bility of a spec­i­fied target behavior and/or to reduce the episodic severity of that behavior. A Behavior Intervention Plan that incor­po­rates con­tin­gent restraint must a) incor­po­rate rein­force­ment based pro­ce­dures, b) be based on a func­tional behavior assess­ment, c) be eval­u­ated by objec­tive out­come data, and d) be con­sis­tent with the sci­en­tific lit­er­a­ture and cur­rent best prac­tices.

Max Benson DIED. Restraints can KILL. And yet ABAI’s posi­tion state­ment is that it is fine so long as it is part of an inte­grated effort to reduce the prob­a­bility of a target behav­iour.

What kind of behav­iour? Any?

ABAI’s posi­tion state­ment is more of the same vague wishy-washy non­sense you find in the BACB’s com­pli­ance code — as long as you include rein­force­ment and have a Plan and it is Necessary and it’s con­sis­tent with Best Practices but we won’t say what Best Practices are…

“Less is better” isn’t enough.

“Consistent with sci­en­tific find­ings” isn’t enough.

Anti-vaxxers have research arti­cles to point to jus­tify what they do. Something being effec­tive doesn’t mean that it is healthy for devel­op­ment or self-esteem.

Once again, too much is left to the dis­cre­tion of the ana­lyst. Trust them to know what the safest restraint methods are, trust them to know when it is jus­ti­fi­able to restrain a human being…

In rela­tion to the treat­ment of indi­vid­uals, Millman wrote: “As long as the aver­sive pro­ce­dure is effec­tive and accom­pa­nied with training and super­vi­sion, under the ABA model you could hypo­thet­i­cally do any­thing.” To clarify this mis­con­cep­tion, sim­ilar to the Certification Council of Professional Dog Trainers (CCPDT) stan­dards of Practice and Code of Ethics Policy, our code also abstains from rep­re­senting training and behav­ioral infor­ma­tion as sci­en­tific, unless it is derived from peer-reviewed and pub­lished research. Specifically, sub­sec­tion 1.01 in the BACB code of ethics spec­i­fies that behavior ana­lysts must rely on pro­fes­sion­ally derived knowl­edge based on sci­ence and behavior analysis when making sci­en­tific or pro­fes­sional judg­ments in human ser­vice pro­vi­sion, or when engaging in schol­arly or pro­fes­sional endeavors. Similar to the CCPDT, our treat­ments must be based on sci­en­tific find­ings.

Scientific find­ings about what? And by whom? Because peer-reviewed studies on small sample sizes or single case design, and per­formed by behavior ana­lysts who do not declare con­flict of interest, is at odds with the rest of the sci­en­tific com­mu­nity and with the autistic com­mu­nity. Further, these studies are based on mea­sure­ments of behavior, not of emo­tional impact of the antecedents on those behav­iors. 

What does the sci­ence have to prove?

Someone could hold up a study showing that denying a child their iPad improves their behav­iour and that would back them up, right?

Science-based. It works. It is proven.

Watson’s Little Albert exper­i­ment was very suc­cessful sci­en­tif­i­cally speaking. And yet it has become a by-word in sci­ence ethics, right up there with Milgrim.

Are your BCBAs required to have doc­tor­ates in the phi­los­ophy of ethics, that you allow them to make choices with such eth­ical risks without strict guide­lines?

Why else would you leave the choices up to them?

A body of research proving that pun­ishing a behav­iour reduces the behav­iour won’t ensure that the child is not psy­cho­log­i­cally harmed. It won’t prove that it is an eth­ical choice.

It doesn’t ensure that the child’s best inter­ests have been con­sid­ered in this par­tic­ular cir­cum­stance.

As far as I can tell there is very little sci­ence regarding the emo­tional impact of ABA on the learner. All of the sci­ence is focused on the target behav­iours.

Guess what? Practising behav­iourism on an organism changes their behav­iour! This is not news! You don’t need a ton of studies on this! We’ve known this for a hun­dred years!

I con­test that many ABA prac­tices specif­i­cally defy sci­en­tific find­ings.

For example, studies of autistic adults shows that “camoflaging” (behaving in such a way as to hide our autism, such as making delib­erate eye con­tact, sup­pressing stims, etc.) is strongly cor­re­lated with sui­ci­dality.

And yet eye con­tact and stim sup­pres­sion com­monly appear as goals in ABA. Dr Zane, you, your­self have written about the impor­tance of teaching imi­ta­tion to autistic chil­dren.

Imitation is the foun­da­tion of many dif­ferent kinds of skills, is an exten­sion of the social respon­sive­ness response class, and is also a social nav­i­ga­tion skill once it becomes a response class (Kleeberger & Mirenda, 2010). In other words, imi­ta­tion can be used to problem-solve; a stu­dent unsure of what to do can refer to peers and follow their lead. However, the pri­mary impor­tance of imi­ta­tion training is that it allows for the teaching of impor­tant skills including toy play, func­tional use of objects, and crit­ical social skills.

Weiss and Zane, 2010

But now we know that imi­tating non-autistic behav­iour is psy­cho­log­i­cally dam­aging in the long term and leads to increased sui­cide risk.

And don’t try to tell me this is pseu­do­science.

Simon Baron-Cohen him­self was involved in sev­eral of the studies which found this. It led to him recanting much of what he had pre­vi­ously said about autism and he now tweets about the impor­tance of autism accep­tance because he real­ized people were dying.

Camouflaging sig­nif­i­cantly pre­dicted sui­ci­dality in the ASC group, after con­trol­ling for age, sex, pres­ence of at least one devel­op­mental con­di­tion, depres­sion, anx­iety, employ­ment, and sat­is­fac­tion with living arrange­ments.…. Camouflaging also explained sig­nif­i­cant addi­tional vari­ance in sui­ci­dality above depres­sion or anx­iety, sug­gesting that the asso­ci­a­tion with sui­ci­dality is, at least in part, inde­pen­dent of mental health.  

Risk markers for sui­ci­dality in autistic adults Ahmad Abu-AkelCarrie AllisonSimon Baron-Cohen & Dietmar Heinke 

“In the short term, cam­ou­flaging results in extreme exhaus­tion and anx­iety; although the aims of cam­ou­flaging are often achieved, in the long-term there are also severe neg­a­tive con­se­quences affecting indi­vid­uals’ mental health, self-perception, and access to sup­port.” 

Laura Hull, K. V. Petrides, Carrie Allison, Paula Smith, Simon Baron-Cohen,Meng-Chuan Lai,William Mandy

So, since there is sci­en­tific evi­dence that forcing autistic people to mask their autism by sup­pressing stims, making eye con­tact, and engaging in scripted con­ver­sa­tion increases risk of sui­ci­dality, does that mean that those things are pro­hib­ited in ABA?

Are ana­lysts every­where expressing horror at the real­iza­tion that they have been pro­moting psy­cho­log­i­cally dam­aging behav­iour for years?

Will ABA ther­a­pists lose their good standing if they engage in behav­iour pro­grams which aim to build NT social skills, thus teaching and con­di­tioning the child to cam­ou­flage?



And what about learned help­less­ness?

It is very easy for a desen­si­ti­za­tion pro­gram to stray into learned help­less­ness, for example.

Behaviourally, it would be dif­fi­cult to tell the dif­fer­ence between learned help­less­ness and habit­u­a­tion and desen­si­ti­za­tion.

If a child is repeat­edly exposed to a noise, and over time dis­plays fewer behav­ioural responses, then is that due to learned help­less­ness or desen­si­ti­za­tion?

It’s an impor­tant ques­tion, because there is sci­en­tific evi­dence that autistic people do not habit­uate or desen­si­tize to cer­tain kinds of stimuli, such as audi­tory stimuli, the way non-autistic people do.


If repeat­edly exposing a child to a noise does not result in habit­u­a­tion, you could easily stray into learned help­less­ness, where the child no longer attempts to escape the noise but con­tinues to expe­ri­ence dis­com­fort.

Learned help­less­ness is a sci­en­tif­i­cally valid method of get­ting com­pli­ance, such that it is used in tor­ture to extract infor­ma­tion from spies.

So… is desen­si­ti­za­tion with regards to audi­tory imput con­sid­ered unac­cept­able by the BACB? Is it con­sid­ered wrong to use extinc­tion due to the risk of learned help­less­ness?

Is learned help­less­ness dis­cour­aged at all?

Even if it is, what if the BCBA run­ning the exper­i­ment hasn’t read this par­tic­ular article on habit­u­a­tion, or others regarding the mal­for­ma­tion of the audi­tory per­cep­tion neu­rons in autistic brains?

How about meta­data studies on autistic chil­dren which indi­cate that the “learning to learn” posi­tion, involving eye con­tact, being seated knee-to-knee with the instructor etc is coun­ter­pro­duc­tive to learning?

What about the fact that ABA still uses Skinner’s book, Verbal Behaviour, fifty years after Noam Chomsky oblit­er­ated it in his scathing review, thus inauger­ating cog­ni­tive rev­o­lu­tion?

You see, ulti­mately, we cannot agree about whether your ethics code pro­tects its learners because we do not share the same pri­or­i­ties, and there is plenty of sci­ence to back up both of our points of view.

In your eyes, if a child learns to make eye con­tact and speak, then the pro­gram has been suc­cessful. You can graph it and pub­lish it and declare it sci­en­tif­i­cally valid.

In my eyes, it may have been abu­sive and harmful.

We can both be right.

Every ABA research paper I have read indi­cates that ABA is proven to work because it achieves its goals.

What isn’t proven is whether its goals are worth attaining, or at what cost.

I mean, shooting someone in the head is a sci­en­tif­i­cally valid and proven strategy… if your target goal is a dead body.

The goals matter.

The validity of the goal, not just the effi­cacy of the tool, is in ques­tion here, as well as the length to which you are willing to go to achieve those goals.

On this note, Millman’s com­ment, “I wouldn’t treat a dog that way.” is a con­flicting point, con­sid­ering the fact that there are sev­eral arti­cles pub­lished in peer-reviewed jour­nals that uti­lize ABA tech­niques focusing on dog training that aligns with the CCPDT’s guide­lines. The CCPDT’s policy also lays out the “Humane Hierarchy”, which advises the methods with which to start. Though not specif­i­cally out­lined, ABA has a sim­ilar pro­ce­dure, just not detailed in spe­cific steps. Like your policy, our code spec­i­fies that we address all pos­sible methods until spe­cific methods (such as pun­ish­ment) are only accept­able as a last resort.

It’s disin­gen­uous to pre­tend that I was crit­i­cizing the use of operant con­di­tioning (what you call “ABA tech­niques” as if ABA came first) when I stated mul­tiple times in my article that I use operant con­di­tioning daily in my work.

My problem is not that you employ operant con­di­tioning. My problem is how you use it, and how you reg­u­late it. Or don’t reg­u­late it, to be more accu­rate.

I was crit­i­cizing the types of goals that are fre­quently chosen for ABA, the narrow-mindedness of using operant con­di­tioning out­side of a holistic con­text given decades of research on child psy­chology and devel­op­ment since the rise and fall of behav­iourism, and the poorly written ethics code which per­mits (and has jus­ti­fied in court) both phys­ical and emo­tional abuse.

So far your letter has done nothing to counter my claims.

You have not demon­strated any effort on behalf of the ABA field to ensure that stim­ming is per­mitted, that planned ignoring is for­bidden, that stan­dard goals and pro­to­cols have been changed to avoid cam­ou­flaging, or that ABA ther­a­pists who employ abu­sive prac­tices such as iso­la­tion or phys­ical pun­ish­ment are dis­ci­plined and made exam­ples of by the cer­ti­fying body.

Millman stated con­cern that our field does not take into account the emo­tions of indi­vid­uals. We dis­agree.  The well-being of our clients per­me­ates our eth­ical code. Our field does care. For instance, treat­ments in ABA target behav­iors that max­i­mize rein­force­ment and inde­pen­dence, both of which sup­port emo­tional health. We do not deny that emo­tions occur.  ABA does, in fact, care about the well­being of our clients. Millman wrote, “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.” Our treat­ments are based on the client’s needs. That is, par­ents may request that this be part of the inter­ven­tion, but if self-stimulatory behavior is not a con­cern of the client’s care­giver and/or does not impede the client’s learning, it is not a focus of their treat­ment.

“We don’t inter­fere with it as long as the parent doesn’t mind it, and it doesn’t get in our way,” is a VERY dif­ferent thing from “We refuse to do it even if the parent requests it because we are mindful of the research demon­strating that stim­ming is emo­tion­ally ben­e­fi­cial and impor­tant to autistic well­being.”

Surely you can see that?

I don’t think it’s enough to say that you don’t abuse the child unless the parent requests it.

I don’t think it’s enough to say that you don’t inter­fere with the child’s coping mech­a­nisms unless it inter­feres with your goals — goals which them­selves are ques­tion­able and may lead to the child com­mit­ting sui­cide some day.

I don’t think it’s enough to say that you would only teach the child cam­ou­flaging at the cost of their mental health if it max­i­mizes rein­force­ment and inde­pen­dence.

All of these caveats are just excuses.

I would refuse to stop a dog from dig­ging entirely. I would tell the clients to pro­vide the dog with an appro­priate dig­ging loca­tion.

I could not give the dog elec­tric shocks for doing it, because before I could, I’d have to refer to the dog to someone else — someone more skilled.

It says so in my ethics code.

While e‑collars are there, at the bottom of the list, they come after refer­ring the dog to other pro­fes­sionals, so how could I ever actu­ally do it?

It is not openly banned but it is made vir­tu­ally impos­sible to do.

Which I like.

I also vol­un­tarily sought cer­ti­fi­ca­tion through the BCSPCA’s AnimalKind pro­gram which does out­right pro­hibit me from using shock col­lars and I think that’s just great because I wouldn’t anyway.

In a decade of training I have never had to shock a dog to achieve my goals.

Part of that is because I am skilled at my job. The other part of that is that I con­sider very few goals worth hurting an animal in order to achieve.

In some instances, such as hand flap­ping, it’s not impor­tant to address stim­ming, but some cases of stim­ming may be harmful or inter­fere with a person’s daily living, how­ever. For instance, behav­iors such as head banging, picking at the skin until it bleeds, pinching, or hit­ting one­self are instances of behav­iors that require imme­diate inter­ven­tion.

Tell me that by “it’s not impor­tant to address stim­ming” you meant to say “it is impor­tant not to impede stim­ming.”

Because those are very dif­ferent things.

If you care so much about the well­being of your learner, why would you ever inter­fere with some­thing like flap­ping? That’s like pun­ishing smiling or crying.

Why aren’t there spe­cific codes against inter­fering against behav­iours that are not only not harmful but impor­tant, com­mu­nica­tive, and ben­e­fi­cial?

As for harmful behav­iours, by all means inter­vene, although in my expe­ri­ence sit­u­a­tions like that are most easily dealt with out­side the realm of behav­iourism.

Introducing harmful behav­iours to this dis­cus­sion is a straw man argu­ment because at no point in my article did I say that I would allow a dog to pluck its tail bald.

If someone wrote an article crit­i­cizing dog trainers for dis­cour­aging dig­ging behav­iour, I wouldn’t storm in and say, “But what if they’re plucking their tail fur?”

(Not the least because they’d wisely answer that the dog might not be plucking its tail bald to begin with if it had been allowed to damn well dig.)

Nevertheless, I should point out that I would also never punish a dog for plucking its tail.

This behav­iour is usu­ally a result of either OCD, in which case pun­ish­ment is inap­pro­priate, lack of enrich­ment, in which case pun­ish­ment is inap­pro­priate, neu­ro­log­ical prob­lems, in which case pun­ish­ment is inap­pro­priate, and so on.

Punishment, in the animal world, is con­sid­ered use­less and harmful when it comes to dealing with self-harming behav­iours. I wonder why it isn’t with humans?

If stim­ming is part of the treat­ment, then behavior ana­lysts will iden­tify alter­na­tive means to find a replace­ment behavior. We agree with Millman that there are times that chil­dren may func­tion better if they are allowed to stim.

“May func­tion better?” Jesus H. Merciful. They’re not machines. They’re human beings. For heav­en’s sake, speak about them like they are people.

Here are some things you could have said instead of “func­tion better”:

  • “Be hap­pier”
  • “Be more ful­filled”
  • “Have better quality of life”
  • “Feel more relaxed”
  • “suffer less anx­iety”

You don’t deny that people have emo­tions but you refuse to refer to them when dis­cussing treating them.

That’s a problem.

You need to remember the mental state of your patients. You cannot prac­tise out­dated anti-mentalism behav­iourism and still talk about sci­ence.

It’s like using Galen as a med­ical text.

Science has moved on from pure behav­iourism. It now belongs on the shelf with many many other tools and should used only in con­text with other aspects of child psy­chology and cog­ni­tive sci­ence.

And you’re hearing that from someone who employs behav­iourism daily in her work.

It is our hope that our response is useful in clar­i­fying some aspects of behavior analysis that Millman seems to mis­un­der­stand. We sug­gest that her view of behavior analysis is out­dated. The eth­ical guide­lines within ABA and CCPDT have strik­ingly sim­ilar guide­lines. It is our hope that these clar­i­fi­ca­tions on our code allow one to see the sim­i­lar­i­ties between our fields, with the pri­mary focus of the well­being of our respec­tive clients and hope that this article cleared up mis­con­cep­tions about the field of Applied Behavior Analysis.


Bacb.com. (2019). Professional and Ethical Compliance Code for Behavior Analysts. [online] Available at: https://bacb.com/ethics/.

Bailey, J., & Burch, M. (2016). Ethics for Behavior Analysts, 3rd Edition. Florence: Taylor and Francis.

Carol Millman. (2019, April 9). Re: Is ABA Really “Dog Training for Children”?  A Professional Dog Trainer Weighs In. [Web log com­ment]. Retrieved from https://neuroclastic.com.

Cooper, J. O., Heron, T. E., & Heward W. L. (2007). Applied behavior analysis (2nd ed.). Upper

Saddle River, NJ: Pearson/Merrill Prentice Hall.

Millman, C. (2019, March 27). Is ABA Really “Dog Training for Children”?  A Professional Dog Trainer Weighs In. [Web log post]. Retrieved from https://neuroclastic.com.

Skinner, B. F. (1983). A matter of con­se­quences. New York: Knopf.

Van Houten, R., Axelrod, S., Bailey, J. S., Favell, J. E., Foxx, R. M., Iwata, B. A., & Lovaas, O.

I. (1988). The right to effec­tive behav­ioral treat­ment. Journal of Applied Behavior Analysis 21(4), 381–384.

No, I’m afraid you haven’t really cleared any­thing up.

You’ve only demon­strated your com­plete lack of under­standing of the nature of my con­cerns with ABA.

If indeed you have moved on from the 1960s as you claim, then modern text­books should con­tain up-to-date infor­ma­tion on men­talism, cog­ni­tive psy­chology, and emo­tional states.

They don’t.

ABA tech­niques when applied to chil­dren should include require­ments regarding respecting the mental health of the child according to the latest research.

They don’t.

Your ethics guide­lines should clearly and unequiv­o­cally pro­hibit the use of cor­poral pun­ish­ment of any type, the use of any known tech­nique known to cause phys­ical and psy­cho­log­ical harm including restraint, iso­la­tion, loss of adult atten­tion etc for any reason.

They don’t.

Your ethics guide­lines should clearly and unequiv­o­cally pro­hibit applying operant con­di­tioning onto reflexive behav­iours such as tics and stims.

They don’t.

It’s not enough to say that ABA can be ben­e­fi­cial when it is prac­tised appro­pri­ately.

Because what I want to know is why it is pos­sible and per­mis­sible to prac­tise it inap­pro­pri­ately?

Why isn’t ABA rec­og­nizing the research on the psy­cho­log­ical harm caused by cam­ou­flaging, the exis­tence of motor chal­lenges which may result in the client being unable to com­plete the goals set by the ther­a­pist, and why is it insisting on author­i­tarian tech­niques which have long since fallen by the way­side in child psy­chology due to sci­en­tific con­sensus?

I’m sorry to dis­ap­point you, but I am still firmly con­vinced that ABA sets goals which do not take into account the emo­tional well­being of the recip­ient, do not account for research on child psy­chology regarding the uses of didactic and author­i­tarian par­enting tech­niques, and ignores cog­ni­tive sci­ence devel­oped since the hey-day of behav­iourism.

I don’t believe that ABA ther­a­pists set out to abuse their clients.

But they do and it hap­pens daily.

If you truly want that to change, then stop arguing with autistic dog trainers and start writing let­ters to the powers that be.


  1. Aaaaaand the BCBAs just fur­ther “rein­forced” that ABA is rotten at its core.

    Great work as always, Carol.

  2. I am a cer­ti­fied dog trainer. I am also likely autistic and have a 3 yo son who def­i­nitely is. I did ABA with abused dogs and my own dog early on not knowing better. It is rotten, was so hor­rible on my dog emo­tion­ally I could vis­ibly see it. I left that, promised to never do that again or work with that trainer. Then I went to a trainer who incor­po­rated the emo­tional well being of the dog into her training. Then I found another trainer’s mate­rial who said it’s the source of all your training. Then I got a dog who we had to reward when he just sat in the group circle well because all he wanted to do was play (adhd dog prob­ably if that exists) .… aba no Bueno for sure. I always ask myself does this allow pro­cessing, swim­ming, breaks, etc. for me, my son and my dogs in all we do.

    And I appre­ciate the response, showing research and expla­na­tion.

  3. Well done!

  4. Psychiatrists and psy­chol­o­gists have no interest in *under­standing* how we think; their only pur­pose is to manip­u­late, con­trol, and enforce con­for­mity with neu­rotyp­i­cals in the way we think. It is destruc­tive, abu­sive, impos­sible, and UNNECESSARY, because we are not “evil” or “defec­tive”, we are simply not like them. Trying to force us to con­form is why our sui­cide rate is so high, but they don’t give a damn about that. They would rather see us dead than allow us to be dif­ferent from them.
    This is the man­date neu­rotyp­i­cals give us: “CONFORM OR DIE.”

    This is why I have been screaming at the top of my lungs my whole life, and will con­tinue to scream it until someone lis­tens:
    AUTISTICS SHOULD NOT BE TRYING TO INTEGRATE INTO NEUROTYPICAL SOCIETY AT ALL. We are many. We know what we are. We know who we are. Our focus should be on coming together amongst our­selves and forging a society of our own where we can be our­selves. A country-within-a-country. Our “Own Private Idaho”. Not nec­es­sarily in a geo­graphic sense.
    Neurotypicals are never going to help us, because *they don’t want to help us, merely con­trol us. We need to learn how to help each other, or we are doomed.

    RISE UP!!!
    RISE UP!!!
    RISE UP!!!

Talk to us... what are you thinking?