A version of this post was originally posted at 37centigrade.org.

By Dusty Jones

If you have a kid with autism, chances are you’ve heard of something called ABA. You may not be quite sure what those letters mean, but you’ve probably heard that this ABA, whatever it is, has been declared by whoever declares these things to be the best and most effective treatment for autism. That might be true, depending, of course, on your definition of treatment. Theoretically, of the many therapy options when it comes to kids with autism, ABA, which stands for Applied Behavior Analysis, offers the greatest possibility that good things will happen for your autistic child as a result of the therapy. That’s the claim, anyway. Unfortunately, the reality is not so cut and dried.

Do We Really Know What Autism Is?

Nobody knows what autism is. Nobody has it figured out yet. We’re pretty sure it has something to do with the brain, but we don’t know what causes it, what makes it worse or less worse, or what can be done about it. Psychology is lost because psychology is a philosophy based on the socially-constructed mind. Kids with autism don’t tend to acquire a socially-constructed mind. The same goes for psychiatry. Psychiatry tries to use chemicals to restructure the mechanisms of the brain so a person complies with social rules. Same with sensory diets, augmentative communication systems, many forms of speech and language therapy, occupational therapy, and so many other approaches, each one offering its own ineffectual or incomplete plan to gain control over a difficult child.

After all the remedies have been applied and explanations given, parents and teachers of children with autism are left with a little creature that behaves in odd and unpredictable ways—ways that defy interpretation.  Perseverative routines for hours. Tantrums that come from nowhere. It’s hard to go to the grocery store when something as random as a beeping cash register can cause an autistic child to have a meltdown, as they say. How do you give a kid a haircut, how do you bring sharp scissors close to his head when he might freak out for being touched behind the ear? These challenges disrupt the most mundane aspects of life. Whatever the chemical makeup of the child’s brain or the psychological interpretation of his behavior or the idiosyncratic sensory input he requires, what’s needed is a solution to the behaviors and habits that disrupt life and make it dysfunctional. Somehow, you need to get it under control just so you can get through a day.

Differential Reinforcement of Other Behavior

Part of the reality is that kids on the autism spectrum often do things that are difficult for their parents and teachers to understand, often disrupting the environment for others. They do things that look weird, or seem aggressive, sometimes even self-injurious. The standard ABA approach to problems like these is to create some kind of differential reinforcement procedure to try to replace the problematic behavior with something else. Such procedures go by different names, but here is the basic idea:

  1. Identify the behaviors that are problematic.
  2. Remove all incentives for engaging in those behaviors.
  3. Provide incentives for engaging in some other behavior instead.

This is called the Differential Reinforcement of Other behavior, or DRO.  My guess is that DRO is the most common procedure used by behavior analysts, because it is the easiest way for many people to conceptualize how behavior changes. If asked the question, “What causes behavior?” many behavior analysts will say that the consequences that result from a behavior cause that behavior to occur more or less often in the future. This is technically true. When applied in certain ways, reinforcement, extinction and punishment are behavioral processes that will, in fact, cause changes in the rate of the behaviors they’re designed to influence. But differential outcomes are just one of many variables that affect behavior. The fact that so many behavior analysts treat reinforcement as a tool to control the behavior of another person speaks to their understanding of the purpose of behavior analysis. Behavior analysis is the science of human behavior. It’s not a program or a set of procedures—it is the scientific investigation into the causes of behavior. Thus, while behavior analysis is often used to control the behavior of autistic children, instead, we can use it to make sense of what they do and help them do it better. When the applied part of applied behavior analysis means nothing more than better behavioral control, I find it very unfortunate for those who are served by ABA.

Another Way of Looking at Behavior Intervention

The approach I have learned to use to understand and improve behavior gets a lot of resistance from people working for better behavioral control. My approach seems wrong to anyone who views behavior in terms of appropriateness and oughtness, rightness and wrongness. The approach does not call for withholding the thing the child needs when he behaves wrongly, but providing it to him at just that moment. For most people, this would constitute a violation of the standard DRO procedure. It calls for a direct response and supportive actions to be carried out at the moment the child begins to engage in problematic behavior. It is an approach that acknowledges the adaptive nature of their so-called problematic behavior and acknowledges that it is only problematic in that it is difficult for their caregivers to manage. The approach described here is simple: Figure out what the child wants and help him get it.

For my colleagues, it comes across as a recipe for making things worse because it looks like you’re reinforcing bad behavior. This is understandable. It does look like you’re creating an incentive for a child to behave wrongly. But this assumes that the actions of another person—whether you praise or don’t praise, give them a reward or don’t give them a reward—is more important to the child than getting his problem solved by whatever means solves it quickest. In my experience, that assumption is baseless. Many of these kids are more interested in spending their time on their own terms than they are in your praise and rewards. Also, in my experience, this type of approach doesn’t reinforce bad behavior; in most cases, it actually tends to bring about a quick reduction of it.

None of this is intended to minimize the significance of challenging behavior. We still identify the behaviors that are problematic and put procedures in place to manage them, but we don’t remove incentives when a child does them. To remove the incentive is referred to as putting the behavior on extinction, and this puts the autistic child in a very difficult position. Whatever problems those behaviors cause for the child’s caregivers, they serve a purpose for the child. The incentives gained from them are more than mere rewards and tokens. They are things like attention from caregivers or escape from an unpleasant situation. For children with autism, escape from an unpleasant situation often requires assistance from someone—they have to get your attention somehow as a first step to getting assistance from you.

Children on the Autism Spectrum Adapt Their Behavior in Order to Protect Themselves

Those problematic behaviors tend to be the means by which a child with autism protects himself from perceived threats and painful circumstances. The world can be a lonely place for a child who learns and lives through the lens of autism, so attention is not just an appetite to fill, but a critical form of support in an otherwise isolated world. Everyday stimuli can trigger frightening reactions in these children, so they learn to protect themselves: they fight against those stimuli, manipulate them, run away from them, or engage in hours and hours of repetitive behavior, all to keep those stimuli far, far away. But their protective behavior is uncoordinated, just as you would expect it to be for someone with a significant learning disability.  Their behavior often puts them into danger they don’t understand.

Whether they make sense to us or not, the child’s attempts at self-protection are at the root of these problematic behaviors. Forbidding a child to engage in them may solve a problem for us, his caregivers, but it worsens the situation for him. It leaves him without the means of protection he’s accustomed to using in an often scary and confusing world.

Under the conventions of orthodox ABA, children learn to throw tantrums, hit others, kick, punch, spit, run away, et cetera, because there is a payoff for doing these things. If you can identify that payoff, you can take it away when they do the bad thing and give it to them when they do the right thing. When using the standard procedures, emotional reactions are regarded as part of the bad thing, so they are also ignored (unless the child becomes aggressive, in which case restraints are often used). Making matters worse, extinction procedures are notorious for inducing adverse emotional reactions as well as dangerous increases in the behavior it is meant to extinguish, and yet this strategy remains ubiquitous in routine behavior analytic intervention.

There is, of course, a rationale behind this part of the process: Once the student begins to encounter reinforcement for appropriate responses, he will “habituate” to any “conditioned aversive stimuli” and the “emotional responding” will simply fade away. Soldiers who come home from war with post-traumatic stress disorder are treated with a carefully-designed systematic desensitization program in which the slamming doors and backfiring car engines that remind them of exploding bombs and death are slowly and gently faded in as the soldier slowly learns to respond and tolerate them. Using the standard DRO procedure for an autistic child is more like chaining the soldier to a bed and firing a gun next to his head until he stops convulsing.

Emotional Effects Cannot Be Separated from the Intervention

Consider the standard procedure for something called food refusal for children with autism who eat little either in quantity or in the variety of foods they’re willing to eat. The procedure, as I’ve seen it implemented, is to hold a spoonful of the undesirable food in front of the child’s mouth until she accepts the bite and swallows it. Treatment sessions of this type can last for long periods during which the student is typically held in her seat or trapped behind a table to keep her in place as she struggles to escape. In a treatment session like that, a child will usually hit, kick, spit, cry, scream and so on. There is a great deal of mucous production, salivation, and sweating, flushed skin and rapid breathing. It is an exhausting process for both student and teacher. If ever the student opens her mouth, the food is immediately shoveled in and the spoon used to block her efforts to spit it back out. If she swallows the food and keeps it down for a predetermined period of time (about 30 seconds) without vomiting it out or just holding it in her mouth, she is given a small piece of a preferred food item and praised. This, presumably, is the reinforcement that increases the likelihood that the student will accept the next bite faster, which will soon be presented the same way as the last.

On those occasions when the child manages to keep a bite down and the preferred edible item is given, there is a pause before the next spoonful is presented. During this period, the physical processes are allowed to return to their at-rest state. Heart rate and breathing slow down. Sweating decreases. Crying and mucous production, flushed skin, body temperature, tensed muscles, everything relaxes for a bit. Then the next trial is started. The spoon is returned to the child’s mouth. The dance of force and resistance begins again. The child fights. Adrenaline once again charges into her bloodstream and prepares her to fight.

In this scene, the hitting, kicking, spitting, the words the child screams at the therapist, these are perhaps examples of learned behavior, coordinated action. But what about the crying, sweating, increased heart rate, flushed skin, salivation, mucous production, and so on? These are clearly respondent, unlearned, reflexive behaviors. These are the behaviors that are tied to emotion. Under the standard procedure, the strength of learned behaviors is assessed and considered before intervention begins. We count the behavior. We know how much it typically happens, how much it happens on the child’s worst day. We consider a margin for error in the treatments we create. We know the process is going to be aversive to the child to some extent, so we try to counteract that with strong incentives for good behavior. We consider the quality and amount of reinforcement. We select items and privileges to compete with whatever the child gets out of doing the wrong thing. But many behavior analysts don’t concern themselves with the physical and emotional effects of their interventions. They don’t assess the level of trauma they are likely to cause, the new “conditioned aversive stimuli” they are likely to create with their aversive control procedures. Why are treatment-induced emotional effects and their potential for damage not assessed along with everything else? These are clearly prevalent in the child’s experience once the intervention begins.

Behavior Is Only Part of the Total Picture of Autism.

One of the criticisms we often receive as behavior analysts is that we oversimplify things, that we do not have a grasp on the complexity of human experience. I appreciate that criticism. I understand where it comes from. But I wouldn’t say it is a limitation of behavior analysis, per se, but more of an inherent feature of the scientific process. Scientists simplify reality in order to make sense of it. We need to be able to change variables one at a time to observe the effects. We run into problems if we change two variables at once. If we change variables that normally cancel each other out, we won’t see the effect of either one. If we change two variables that operate in the same way, we’ll see an effect that is larger than we’d expect. Because we study behavior empirically, which means we study behavior in terms of what we can see and observe, we focus very closely on what we call behavior-environment interactions. These are discrete events in which a behavior causes an immediate effect in the environment. We study behavior in terms of what it causes in a person’s surroundings just after the moment the behavior occurs. That means we tend to view those immediate outcomes of behavior as the key variables in determining whether or when that behavior occurs again. But behavior analysts tend to forget that the simplified units of behavior are simplified for the sake of experimental analysis, and are therefore only part of the total picture, the total experience.

It is often said that children with autism do well following tightly structured routines. This is true. The level of predictability in an established routine is very helpful for the child trying to get through his day. But what if the same aversive event occurs reliably at some point in the predictable routine? Let’s say that each day after lunch, the child goes to art class, and the smell of the paint reminds him of a time when he made a mess with paint and got a harsh scolding. From his perspective, the paint is potentially dangerous because it has been associated with being scolded, so the room that has that paint smell is potentially dangerous, too. Knowing he’ll go to that room after lunch every day gives him something to worry about during lunch. If he refuses to eat or “has a meltdown” when he’s supposed to clear his tray, a tight focus on behavior-environment interactions will see his caregivers run to him to get it under control. After observing this phenomenon a few times, the conclusion will be that he does it for attention, and the new written procedure will instruct his caregivers to ignore him when he refuses to eat or clear his tray. Such a focus won’t reveal that it’s actually the next item on the schedule—the smell of pain in the art room—that is causing the problem. Every step that leads to the moment that he enters the art room could signal to him that he is approaching that unpleasant moment and cause him to react problematically as he encounters those signals throughout his day. It could start as early as when he wakes up in the morning. Waking from sleep and beginning the day puts him on a path that inevitably leads to that critical aversive moment when he enters the art room. So, aside from those discrete units of behavior-environment interactions, his total experience is one in which he has certain elements (the structured routine) pushing him forward through the day, and other elements pushing him back. He experiences a resistance to move forward because moving forward leads him to that single moment of uncertainty and fear. He is pinched in the place where those forces come together, like two fists pounding against each other. He makes his way through his daily schedule in the context of those opposed forces, like a hungry person with lactose intolerance and all there is to eat is rich, creamy cheese.

What If Making Them Do Things for Their Own Good Isn’t Good at All?

We would like to make things easier for kids caught in such a situation. Perhaps we change the antecedents (the events and stimuli that lead up to the behavior) to try to relax some of the pressure that pushes kids forward. Perhaps we relax the timetable and allow the student to go at his own pace. In doing so, the resistance to the painful moment may become the operative source of motivation. If allowed to avoid it, he may never enter the art room. This can be problematic. Perhaps we can allow him to forego art, but what if similar aversive events occur in the lunch room, or the classroom, or on the playground. If kids go at their own pace, they might avoid these environments altogether and never get to experience the good things that can happen in those places. They miss opportunities to expand their repertoires and enrich their lives. This is often the reason we feel compelled to push them through their resistance and force them into the situation. What they stand to gain, we reason, is worth the pain they have to endure to gain it. After all, a student who refuses food is at risk of malnutrition. The ends justify the means.

But what about the pain we cause using an approach like this? What of the trauma we cause? If we treat food refusal by forcing children to eat in spite of their emotional experience, aren’t we taking the risk that we are teaching them to eat out of a fear that if they don’t, they will be put through another “session” and traumatized some more? Have we really helped them if this is what we cause? Or are we adding to the total experience of trauma and duress they experience every day living in a world that doesn’t make sense to them? This world isn’t designed for comfort for any of us. How much less for a person with a profound learning disability struggling to process it and navigate it?

I believe that to be the most important consideration in helping a child with autism—that every single external stimulus hitting their receptors could potentially bring pain and trauma. The sensation of heat radiating through a car window under the glare of the midday summer sun is nothing for the rest of us to fear. We have learned to identify the source of the heat. We’ve learned that cold is its opposite, and that the cool air blowing on us from the dashboard is designed to compete with that sunny window heat. We know how to pull the visor down to get a little shade. But for a child with autism, such learning may not take place automatically. Having a learning disability doesn’t just mean you’re unable to learn what they teach in school. It means you don’t learn how other people learn in possibly every aspect of life. Most of what humans do in this world they learn to do. So, the child with autism may have no basis for understanding why he is unbearably hot from the sun glaring through the car window or how to remedy the problem. Yet, I have never seen a teaching program designed to teach a child to take control of a car’s A/C. In fact, it is more likely that an autistic child attempting to operate the A/C will be blocked—redirected, as we say. We can’t let him operate the A/C because he doesn’t know how and he might break it. And so he is left to feel that painful sensation without understanding and without relief. He is left to wonder and worry: What is that? It hurts. It feels like fire. Will it burn me if I stay here? Am I in danger? I am in danger. I’m in DANGER! And that’s where the panic begins, the drive to escape, the drive to protect oneself in whatever way possible. Perhaps that involves removing his seat belt or attempting to escape from the car—no matter that it is rolling down the road at speed. What does that mean to a child with autism? Perhaps his escape even involves attacking the driver or himself.

This is the moment when supportive care should be offered—at the point when the child begins to engage in these forms of self-protective, even violent, behaviors. The standard DRO approach stands back from this idea, if not disallowing it entirely. According to this procedure he child must stop engaging in these extreme behaviors before his situation is improved, because if the situation is improved by means of these behaviors, then they will be reinforced and the child becomes more aggressive, not less.

This is where I question the nature of the response itself.  There are certainly operant processes involved here, processes that have taught the child that aggression is useful at times. There is coordinated action. The child slaps the driver, not the rearview mirror. The child may have slapped the mirror in the past when his behavior could be better characterized as flailing about. But the specific response that caused a change and an improvement in the child’s situation was the aggressive behavior (perhaps the driver pulled the car off the road, perhaps under some shade), and the behavior that preceded that change was the behavior that got captured under the processes of negative reinforcement and escape.

Viewed this way, it should be apparent that the challenging behavior is going to cause an improvement in the situation in spite of our efforts to hold things constant, to keep the situation unpleasant until the child attempts a more appropriate form of escape. Even if the change in the situation is the introduction of other aversive stimulation (such as being physically restrained), the stimuli that triggered the problem in the first place are likely removed or made less salient by the new situation. This change in itself is enough to reinforce the violent behavior that brought it about.

So, why not get ahead of this effect? Why not take the aggressive, problematic behavior as a cue that the child, unable to express his needs with words, has suddenly encountered a situation he doesn’t understand and that exceeds his ability to respond effectively? Why not take the opportunity to help him find comfort and return to a state of contented equilibrium? We can assess the environment where the problem occurred, look for potential sources of pain that led to it, and work towards ways of helping the child function more successfully when he encounters that situation again. But—and this is key—we can only do that if we stop worrying so much about what the student does, and start worrying about what the child needs. That is, what the child needs in that moment to make himself as comfortable as we caregivers are able to do when we encounter similar sources of discomfort, pain, and uncertainty in our own daily experiences.

No Behavior Happens “For No Reason”

The rationale for this approach is based on the notion that any human, learning disabled or not, will engage in flailing about when we perceive ourselves to be under threat or duress. It’s a genetic condition. We are built with the capacity to charge ourselves with adrenalin, enrich our blood with oxygen, enliven our senses and move our thoughts in quick patterns to discriminate the potential means of escape and protection. It is the learned, behavioral component of the fight or flight reflex that evolutionary selection endowed us with that enabled us to survive as a species. We like to think of ourselves as qualitatively different from the animals, as if our sophisticated selves, functioning in this sophisticated social world are above and superior to the animalistic aspects of our natural selves. I don’t think anyone can seriously embrace this notion. We evolved from that hairy beast. What makes us think it’s not still part of us? We all perceive the panic that overwhelms us in uncertain moments, and the drive we feel to find our safety and comfort. It is no different for a child with autism, or a child with a physical disability, or a child with an emotionally charged history, or just a regular so-and-so who has experienced the pain of losing a friend. It’s the same for all of us, and if anyone criticizes us we are ready with our explanations and justifications. For some reason, when kids flail about in ways that seem aggressive, we assume that there is something wrong with them, that their broken neurons make them inherently unpredictable. This is not the case. They are humans in need, and that is the first perspective we should take of them, and perhaps the only perspective we take for a while. Solve that problem first, and the rest of it gets easy.

Learning Is a Natural Process

Finally, a word on the nature of learning. Learning is a natural process. It is something that happens without our help. Sure, we can discover the variables that contribute to and cause a person to learn. We can learn to manipulate those variables to bring about the kind of learning that meets our personal and social objectives. But learning happens most efficiently and effectively when it happens naturally, and I’ve learned that it’s usually best to let the world operate as it does without trying to force it towards some predetermined goal.

It’s funny—such a simple concept, putting our focus not on the outcome but the process. The process of learning means doing something you can’t do or understanding something you don’t understand. To learn, you must do something that is counterintuitive to your current framework of understanding. That is the essential dilemma in learning. That’s what makes it so difficult for kids with autism. They’re sensitive to discordant rhythms and sensations. They value synthesis, resonance. They like it when things line up, when things happen when they’re supposed to happen, and in the familiar way that they normally happen. (“This is the time of day when we throw tantrums, cry, fight, scream and wallow around on the floor for 6 to 10 minutes. After that we have snack.”) So the dilemma in learning, acutely felt by kids with autism, but true for all of us on varying levels of intensity, is that learning is psychologically dissonant and behaviorally effortful.

To work against the dissonance there has to be trust. You have to be able to trust that the thing you don’t understand will make sense if you make the leap. And for a child with autism, this usually means trusting another person… which is problematic when angry faces and punitive reactions to the child’s behavior make other people untrustworthy from the outset.

The Key Is to Help Kids on the Autism Spectrum Feel Safe and Relaxed

To work against the effort, you have to make the transition as smooth as possible, the steps as small as they can be and still offer forward momentum. This requires gentleness and patience. Neurotypical people can learn in leaps. We can learn several things at once. It’s not as easy for kids with autism. They are very narrowly focused because they are focused first on the potential for unpredictable, random things that come out of nowhere. Unless they’re comfortable, they are always on guard and you can only get them to learn one or two things at a time. But if they are allowed to feel safe and content, if everything happened the way they understand it should happen, and things are less likely to fly out of nowhere and attack them, then they don’t have to be so watchful, so vigilant. They can relax, and when they’re relaxed, you can catch their interest and teach them many things, happily.

That need to feel safe and content in order to learn is a requirement for all of us. We’re not as susceptible to attack by random, everyday stuff, but if we do come under threat, what we tend to learn is how to escape or manage the threat. We don’t really appreciate the content of any other lessons we might learn because we’re more focused on what’s minimally and sufficiently necessary to control the threat at hand. If we feel safe, there is no imperative to escape and we can take an interest in the topic, if it is interesting, and learn for the enjoyment and pleasure in it. That’s the kind of learning that enriches life. That’s the kind of learning that builds things together, like skills and abilities, understanding, confidence and wisdom.

Photo Credit: Lance Neilson/Flickr

Dusty JonesDusty Jones earned his Bachelor of Science in Applied Behavior Analysis from the University of North Texas in 2001, and a Master of Science in Behavior Analysis at UNT in 2007. He has served various populations, both in the United States and overseas, including individuals with autism and other developmental disabilities, child victims of abuse and neglect, individuals with acquired brain injury, incarcerated military personnel, parents and families of individuals affected by behavioral difficulties, care and educational staff of residential institutions, and special education students in public schools. He uses non-aversive interventions, and specializes in severe challenging behavior, including oppositional and defiant behavior, aggression and self-injury. He has developed educational programs for parents and professionals and has conducted many presentations and workshops on Behavior Analysis covering clinical and social topics from ranging from challenging behavior and early childhood development to human sexuality, culture, and economics. He has co-authored two chapters on self-injurious behavior in the International Handbook of Autism and Pervasive Developmental Disorders, and currently authors educational materials for behavior analysts. See more at 37centigrade.org.