The Great Behaviorism Manifesto
Hell is a place where children go severely underpunished—or severely overpunished. The only problem is I can’t yet tell which.
Framing the debate
I read an academic paper once that found that students learn a great deal from debate—but they walk away believing their side of the debate more strongly than before. You can imagine that a debate in a political science class might deepen students’ political ardor, but not leave them much more informed—as just one example. Debating the law of gravity might be an interesting learning activity, or it could leave students with misconceptions and misunderstandings.
The key take away from the article was simple: if you are going to debate a topic in classrooms, make sure to follow up and clarify what we know, what we don’t, and when appropriate, what the scientific research says on a given topic.
I’m doing that here. Each semester in my educational psychology course, I teach students about behaviorism by asking them to have a debate on the topic—the great behaviorism debate of (insert semester date here). Well, it’s summer of 2023, and I’d like to make my summation of the matter a little more clear—so that my students leave with their own opinions on behaviorism, but also with a clear sense of what I believe on the matter.
(By the way, I have not cited the above paper because I couldn’t find it—if there is an enterprising young student looking for a few point of extra credit, a link would be appreciated!)
The Question
In class, I have posed a simple question: are punishments and rewards right to use with children? Particularly in the classroom?
The tension here is profound. I am sorry to have to leave the question at your doorstep, because the longer you study it, the more necessary it becomes to be conflicted. You will see things and hear things that you will never forget, despite your trying, and so I urge you to be wise. But you need to understand the profound immorality on either side of the debate.
If you’d like a still-haunting, but relatively innocuous example, see below. But consider yourself warned. You will not soon forget his screams, should you choose to click.
https://www.theguardian.com/us-news/2020/mar/05/us-bans-electric-shock-treatment-children-boston-area
If you don’t click, there is no judgment from me. I only want to instill in you the depth of the debate, and how serious it is. I do not pretend to be able to answer such things clearly and completely, but I think I can offer some thoughts that might be of use.
If nothing else, I have reached a profound state of conflictedness, and that seems to me to mean that I have at least appropriately grappled with the question.
More than Carrots and Sticks
First, a technical note. Our debate on behaviorism isn’t actually about behaviorism, and that’s worth noting in the first place. Rather, our debate is something along the lines of “is it good, moral, and scientifically supported to use rewards and punishments in the classroom?” I need to begin by reminding you that behaviorism is far deeper than merely consequences, or carrots and sticks. If you already know about behavioirsm, feel free to skim—but I don’t want you to leave without understanding what behaviorism is—and how our debate is actually only a tiny slice of a much bigger set of concepts.
When I ask chatGPT about behaviorism, I get a very good summary. Here it is:
Behaviorism is a school of thought in psychology that focuses on the observable behavior of individuals rather than their internal mental processes. It emerged as a dominant approach in the early 20th century and was heavily influenced by the works of psychologists such as Ivan Pavlov, John B. Watson, and B.F. Skinner.
Key principles of behaviorism:
Empirical approach: Behaviorists emphasize the use of empirical methods and experimental research to study behavior. They believe that only observable and measurable behaviors should be the subject of scientific investigation.
Stimulus-response associations: Behaviorists propose that behavior is a result of the association between external stimuli and the responses they evoke. In other words, behaviors are learned through interactions with the environment.
Conditioning: Behaviorists introduced the concept of conditioning, which involves the process of learning through associations. There are two main types of conditioning:
Classical conditioning (Pavlovian conditioning): This occurs when a neutral stimulus becomes associated with an involuntary response through repeated pairings with a stimulus that naturally triggers the response. Famous examples include Pavlov's experiment, where dogs salivated at the sound of a bell after it was paired with food.
Operant conditioning: This involves learning through the consequences of actions. Behaviors that are followed by positive outcomes (rewards) are more likely to be repeated, while behaviors followed by negative outcomes (punishments) are less likely to recur. B.F. Skinner is well-known for his research on operant conditioning using the Skinner box.
Objective approach: Behaviorism rejects the study of subjective experiences, thoughts, and emotions since they cannot be directly observed and measured. Instead, it focuses solely on external, observable behaviors.
Nurture over nature: Behaviorists tend to emphasize the role of the environment in shaping behavior rather than innate or genetic factors.
Behaviorism had a significant impact on psychology and other fields, influencing areas such as education, therapy (behavioral therapy), and animal training. However, it has been criticized for oversimplifying human behavior and neglecting important cognitive and emotional processes. As a result, contemporary psychology incorporates elements from multiple perspectives, including behaviorism, cognitive psychology, and neuroscience, to provide a more comprehensive understanding of human behavior and mental processes.
My primary point here is this: as we debate conditioning tactics in classrooms (rewards and punishments, often characterized as “carrots and sticks”) let’s keep in mind that behaviorism is deeper than merely carrots and sticks—and there’s much to it that is worth appreciating (or debating!) separate from mere consequences.
But I want to be clear, too, that I am, by-and-large, compelled by behaviorism. I want to make sure that my reasons why are clear here and now. These aren’t comprehensive, but they give you a flavor of my priors, and that’s enough for the present task:
Behaviorism is a Defense Against Pseudo-Science. The empirical focus of the behaviorists is such that some went so far as to dismiss anything that happens “in the mind.” The only thing that mattered was behavior—all else was pseudoscience. In fact, that’s an absurd first principle—but it’s a really good framework to build a reliable set of scientific observations. Psychology is plagued with subjectivism. I’ve said before that psychology broadly is in existential crisis over it’s ability to correctly diagnose (what I call the diagnostics crisis). I get off the bus with some of the more radical thinkers, but it seems obvious that a rigorous science will require us to rely only on what is observed objectively.
Behaviorism Squares With My Anecdotal Observations—and Morality. When people argue with behaviorism, they tend to do so on moral grounds. In fact, I find behaviorism’s focus on nurture over nature to be incredibly empowering: we can teach others to be agents, rather than objects to be acted upon. If you are a parent, a teacher, or a manager, you want to know how to motivate people—and behaviorism is concerned with motivation.
Behaviorism Works—Independent of Morality. One of the strongest critiques of BF Skinner is that he treated humans as animals. That critique has never felt right to me, however. I feed my children, make them sleep, and correct bad behavior—is feeding my children treating them like animals? When we get away from the normative (ethical) concerns, and get into positive (scientific) concerns, we will see a lot more value in behaviorism.
For example, Skinner was once asked how he justified using animals in war. His response: “I am opposed to military use of animals in war. I am also opposed to military use of men in war.” What Skinner was able to accomplish, however, using behaviorist techniques was nothing short of remarkable: Training pigeons to guide missiles (which did not end well for the pigeon), training rats to dance on ladders, and teaching pigeons to play ping pont are all incredible, breakthrough accomplishments if you can separate them momentarily from ethical concerns, and address those later. (Though we do have to come back to them later.)
One final note. I know of some people who have a strong distaste for behaviorism or BF Skinner because he is an atheist. I find this borderline monstrous. He is a luminary in the field of psychology, and his research should be evaluated on the extent to which it holds true. There will be plenty of terrible ideas peddled by theists, just as there will be many good ideas peddled by atheists. I will evaluate his ideas on the basis of their integrity—not on whether they disagree with me on other issues. If Hitler prominently advocated for regular showering, it would be a mark of my intellectual immaturity if I went around smelling funny in order to show just how much I disagree with the man.
Similarly, Skinner’s ideas strongly pushed a view that has come to fruition most clearly in the writings of Sam Harris: that we have no free will, and that we are only a summation of influences acting upon us. I find Harris’ ideas to be empirically wrong, morally dangerous, and (here’s my greatest insult) poorly argued. That BF Skinner had some ideas that align with those of Harris (that people are only products of their environment and conditioning, and thereby have no will) then he believes something with which I disagree vehemently. And yet, at the end of the day, Skinner was a good scientist who was right on a great many other things. I simply don’t feel it’s necessary to buy into his radical view of free will to believe that behaviorism has important insights into human motivation.
By the way, if you’re wondering about agency, I suspect that after spending so much time in the lab with mice, he realized that they have little choice indeed. I wonder if he considers whether the experimenter does.
On Punishment
A professor of mine once shared that he had polio when he was young. His father, under the advice of doctors, was told that he needed to extend his legs manually each night, and then let them relax, over and over again—or he would never walk. His father did so, under the wails of his precious baby.
And that, in my opinion, is a critical part of what it means to be a good father—and one that we are losing.
When I was a teacher, a young man that was a behavioral challenge came to me after school. He asked me, “what constitutes child abuse.” He was in seventh grade.
I remember calling CPS, talking with him, and doing everything I could, but those moments never leave you. He later explained some of what he passed through, and that won’t leave me either.
I have a theory that much of our societal social intuitions come from psychological scarring. You can imagine that part of why we are so anti-authority is because of the horrors of the 20th century, including Mao, Stalin, Mussolini, and Hitler. I wonder what our intuitions would be if we had passed through times of chaos and anarchy instead of authoritarianism.
And at the same time, I have had the experience—as I’m sure you have—of interacting with youth who have not been given any meaningful consequences for their actions. Without intending to, these parents raise little monsters who are entitled and undisciplined. Jordan Peterson has a chapter on this in his first book on rules for life that goes something like “don't let your kids make you dislike them.”
I am profoundly concerned that we seem to be either abusive on the one hand or permissive on the other. I should note that I find this to be particularly a problem with special education students, and that too many special education teachers believe they are helping when they are harming by coddling students both academically and behaviorally. The reality is that we are game-playing machines, and making the game too easy is the root of too many dysfunctions.
I am going to detail in a moment what I think the answer is, and it isn’t quite “in the middle.” If there’s a choice between abuse and doing nothing, then let’s just aim for moderation, yes? No—that’s the fallacy of moderation, and it’s a trap. The goal here isn’t centrism, it’s wisdom.
There are some things that will punish a child: gravity, society, the laws of nutrition, and so on. The question is not whether they will get away without punishment—the question is whether that punishment will be given deliberately by someone who loves them, or coldly by someone (or some force) that does not.
When one of my children was younger, he had very emotional trantrums. I would take him to a room and hold him. He would scream and kick and yell, and I would try to keep it together. Inside, I was weeping. It was among the hardest things I’ve ever had to do.
Except that I had the distinct feeling that he was going to have those feelings no matter what I did, and we could choose to help him control them, or allow him to do as he saw fit. I chose the former. It was an investment—hard and sometimes terrible, but absolutely worth it. He still has strong emotion, but I could not be more proud of my little boy: self-posessed, thoughtful, and aware of his feelings.
It is my strong belief that everyone who intends to be a parent needs to butcher a chicken, souture a serious wound, triage patients, or fire a beloved employee. We have lost the emotional durability to do hard, worthy things.
We claim that this is about caring for the child, but I am afraid I know too much psychology for that. I think it’s about us. The immediate emotional reaction is a desire to shrink: to run from the terror of being the judge, jury, and executioner, because it is, in fact, a terrible weight. We run and run, and our children grow without care or discipline. So when we consider the idea of punishment, our elephant—our emotional selves—shrink back. Our rational selves connect the dots—we make up a story that makes us feel at peace.
“Punishment isn’t good for kids.” “I avoid it for for them.” “I don’t want to hurt them.”
So go the cries of the coward—just as the abuser responds “punishment is good for kids,” and “I’m doing this for them,” and “sometimes you have to hurt them.”
Aversion to Aversives
A podcaster I listen to has said that he has sworn off physical punishment for his children because he treats himself as a loaded weapon at all times. I think he is profoundly wise. Please, let me be clear: the evidence on physical punishment—and spanking in particular—is very clear, and very negative.
I also believe that we will be held to account for every tear of our children unjustly shed. I know that Jesus wasn’t talking about this—at least, I didn’t think he was?—when he talked about offending “these little ones,” but I think it still applies. Heavy will be the burden on the day of judgment for those who mistreat their children.
And yet, because of our fear of harming our children, I fear we have let the pendulum swing too far the other way.
I have recently heard of “positive-only parenting,” which strikes me as the hippie-dippiest thing I’ve ever heard. I know of schools who really believe that children simply need to be talked to in order to act right—and they eventually find that not suspending children still results in suspensions, only for far more weighty offenses.
You can choose when to have your battles, but you will have them.
The question is one of ethical aversives. In psychology, we refer to unpleasant stimuli as “aversives,” often used as punishments of some kind: a quick smack of the bum, hand, or head, taking away a toy, or placing a child in time out. It is good and right to ask whether it is moral to administer such things, and how best to do it—but I want to be clear that behind the curtain of every anti-aversive argument is an argument against all aversives.
Get rid of suspensions! They’re bad for kids! (I couldn’t agree more, nothing sends a worse signal than “you were bad, so have three days to play video games without parental supervision.”) But then it’s on to how in-school suspension is a bad idea, and how detention keeps kids from playing with friends, and how recess is a human right (wrong, but also, let kids have recess) and suddenly you realize that there is no line in the sand!
I am not wise enough to tell you where the lines ought to be—but I feel very morally certain that there should be some. I have seen the moral mess that comes from societies (even in miniature—we call them schools) that attempt to have no punishments at all, and I daresay that the adults involved are engaged in acts of cowardice and soul-level negligence. They are not only facilitating and enabling, they are actually growing crops of morally malformed children.
In Praise of Praise
Fortunately, not all of conditioning is punishments: there are also positive reinforcements to consider.
We have established that behaviorism is greater than merely rewards and consequences. I have also explained that I appreciate many of the broad tenets of behaviorism, even if I don’t agree with all of them, and have finished off by explaining that our concerns over behaviorism should limit themselves to whether behaviorism itself has merit—not tangential issues like whether Skinner was an atheist. Now let’s return to our initial question: should we use punishments and rewards in schools?
I think carrots and sticks are useful, and that it is practically impossible to do otherwise. Here are some principles of why I believe in a carefully implemented behaviorism.
Behaviorism works
ABA Therapy (Applied Behavioral Therapy) is a treatment for people with autism that is based strongly in behaviorism. Basically, reward good behavior, punish bad behavior, and do so with some amount of immediacy. ABA is controversial. Ivar Lovaas, the pioneer behind ABA, would originally scream, slap, and use other aversives on children in order to teach them how to act in a way that is acceptable. He also said that with the proper therapy, he could make autistic children “indistinguishable” from their peers.
Now, I don’t want to get into the full debate about ABA here, but it’s worth clarifying that I’m doing what I can to read up on things, even if you don’t agree with my position. I’ve thumbed through a couple dozen sites, but here are a couple that explain things thoroughly.
https://www.totalcareaba.com/autism/aba-therapy-controversy
https://gsep.pepperdine.edu/blog/posts/debunking-7-common-myths-about-aba-therapy.htm
I have only watched a portion of this video, but I feel strongly that I need to represent the views of people with whom I disagree as faithfully as I can, and the portion I did watch was thoughtful and clear. It also speaks from the perspective of people who have undergone ABA therapy, which is a perspective I can’t speak to.
Another source that is worth your time—though it is far lengthier—is “In a Different Key,” by Donvan and Zucker. It carefully navigates the controversy of Lovaas and ABA while measuring both the pain afflicted on children and also the potential upsides for autistic children.
I should also include a credible video that gives my own perspective. Here’s a video that I think gives a good summary from one of the links I shared earlier.
It’s important to understand that the choice is not between children who are treated to harsh rewards and punishments and children who are not. Rather, it is between children who are subject to self-injurious and children who are subjected to therapy—often to serious gains. From the Wikipedia article on Lovaas, (citing a review of the literature of ABA therapy): "Children who receive early intensive behavioral treatment have been shown to make substantial, sustained gains in IQ, language, academic performance, and adaptive behavior as well as some measures of social behavior, and their outcomes have been significantly better than those of children in control groups."
One way to tackle thorny issues like ABA therapy is to take things one step at a time: first, the moral question, then the scientific one. (I am not suggesting that we ignore morality. I am suggesting that we address them very seriously—just at separate times.) The moral question is this: if there is a chance to significantly improve IQ, social functioning, and behavior, is it worth experimentation and therapy? Even if the therapy is harsh and sometimes unpleasant?
Next, the scientific question: in this case, does ABA work? The answer appears to be, yes. That is, it changes behavior, encourages good behavior, discourages bad (or unusual) behavior. Given that ABA therapy is time dependent (it works better at earlier ages) there is some urgency to apply it early and without hesitation.
By the way, since Lovaas’ initial experiments with aversives (unpleasant punishments meant to modify behaviors) various ABA therapy organizations have created codes of ethics, and ABA therapy has gotten away from aversives (including slaps, screams, and so on) because they simply don’t work over the long term.
ABA is only a microcosm of the greater school of thought of behaviorism, and it gives a good inkling of how to best tackle the ethical issues: separate them into morality first, and then science second. If a therapy is beyond-the-pale (or a student consequence in schools) it shouldn’t really matter whether it “works” or not. On the other hand, our moral intuitions against punishment of any kind may be so strong that we do not recognize that potential benefits to students from clear procedures and strong structures.
Positive reinforcement works better
Behaviorism has become nearly synonymous with BF Skinner—but when people think of behaviorism, they ignore one of Skinner’s most famous dicta: “Properly used, positive reinforcement is extremely powerful.” What’s more, many people critique Skinner by using Skinner’s own quotations: “A person who has been punished is not less inclined to behave in a given way; at best, he learns how to avoid punishment.”
Reinforcement requires articulation of values
Part of the genius of positive reinforcement is that it forces you to articulate what you think is good and right.
Imagine with me a teacher who does not believe in dollarbucks and token economies. This teacher believes that such rewards were artificial, and that students should do the right thing because it is right—not out of fear of punishment or desire for reward. Now to this hypothetical teacher (who is not particularly hypothetical) I always have two questions: do you collect your paycheck—or is it enough to simply know that you have done good work in your professional sphere? Second, does this teacher feel equally about punishment as reward? It has been my experience that many teachers feel that it is not their job to reward good behavior—but it is their job to punish bad behavior! (In at least one case, I fear that the teacher simply didn’t want to have to do the hard work of communicating what good behavior looked like!)
The reality is that how to reward and punish behavior may be morally frought—but the obligation to communicate what you desire is not. If you want a better relationship with a loved one, communicate when they do something you appreciate. I don’t think my wife does things because I want her to, or because she desperately needs my praise—rather, the positive reinforcement of praise communicates that I really love that thing she does. Take five minutes each night to let someone know that you are grateful for something they did—and I’d bet dollars to donuts that the rate they do that thing at will increase.
This isn’t my idea—and I’m not even sure it’s Jordan Peterson’s either—but Dr. P explains it in even clearer terms below:
When you communicate to people who love you that you like behavior X, they will do more of X! And not because they want your approval, but because they want you to be happy! That’s what love is, after all: a deep desire for the well-being of another. You’ve now given them a treasure map with a clear path to what they desire: your happiness. (You should try it some time.)
We are motivated by fear
Positive reinforcement is more effective than punishment, as I’ve just explained, but now I need to throw a wrinkle at you. There is one exception to this rule, and that is that we are profoundly motivated by fear. Dr. Jordan Peterson is now a pop-figure as much as a psychologist, but his explanation of fear-based motivation is useful here, and based in the scientific findings on behaviorism—and the uplifting life-coaching at the end is pretty good too.
But I share this with you for a second reason. Not only is positive reinforcement effective compared to other types of conditioning, but punishment is ineffective. Why? Because we are more scared by the fear of the punishment than we are by the punishment itself. Prison may be useful—but very few of us are avoiding prison as much as we are the fear of doing the wrong thing, being caught, or being punished. If you send children to the principal’s office for punishment, or call home, it will only work if it does not become regular—and if the fear succeeds in producing a different set of behaviors.
95% of our motivation is not driven by the actual consequence, but by fear of the consequence. When someone has passed through the consequence and learns that it’s not all that bad, the power of the consequence is gone.
A God of consequences
If you don’t agree with consequences and rewards, that’s just fine—but restorationist scripture (and Latter-day Saint scripture in particular) is filled with a God who is always promising conditional blessings, based upon our faithfulness to covenants. In fact, it has become so clear a part of our theology that I regularly mark scriptures in blue and red: red for commandments, and blue for concommitant blessings. If you don’t like rewards and punishments, you’ll have to argue with Someone above my paygrade.
Rewards and punishments are training wheels
When I was a kid, my dad taught me that we can obey God for three possible reasons: first, because we love him, second, because we want blessings, and third, because we fear negative consequences. I like this teaching—but it is the next part that is most important: God will accept our right behavior no matter what the reason is—but he wants us to obey for the right reasons whenever possible.
Rewards and punishments are good if they teach us to act right. If they become transactional, less so. Once, when I saw my daughter arguing with her siblings, I told her to quit arguing or I would send her to bed. She looked at me, yelled a final angry argument at her brothers, and happily put herself to bed—happy to have made a worthwhile trade.
We have a strong intuition against rewards and punishments, I think, for this very reason: because we fear that anyone acting out of fear of consequence might not be doing it for the right reason—or not for long. The answer to this, however, is not to eliminate all rewards and punishments, but to use them explicitly as a mechanism for teaching proper morality. Instead of “I will give a jelly bean to anyone who does the right thing!” it becomes “I love the way you did the right thing—and I’d bet you’d do it even without the jelly bean, but I want you to know that I’m grateful for it anyway!”
Eliminating rewards is practically impossible
If we were to attempt to eliminate all rewards, we would need to start with making eye-contact with babies. The reality is that we are wired for reward, and the rewards that are most effective are most primal: food, shelter, intimacy, esteem—even base communication.
Effective rewards and punishments require wisdom
This may be the least scientific of all my little notes, here, but I suspect that even the most radical behaviorist would agree: it is not a question of whether rewards and punishments work, but rather, of how expertly we can apply them.
Most of us have a strong aversion—a moral intuition that pushes you away from—too many rewards and punishments. That’s for good reason. You don’t want to have entitled children who only do something for a reward; you shouldn’t punish misbehavior into oblivion rather than teach students how to behave correctly, and then model the behavior. If you are rewarding a three year old for pooping on the potty, you’re doing things about right—but if it’s still happening at age 16, you have bigger problems.
The way to address this is to wean children off of rewards as soon as you can—and teach them about intrinsic motivation. (The reality is that too few of us are willing to extrinsicly motivate our children to intrinsically motivate themselves!)
The example of praise is perhaps most useful here.
According to ChatGPT, there’s a right and a wrong way to give praise, based on Carol Dweck’s book, Mindset. If you already know about Dweck’s basic theory, skip ahead—but here’s a review if you need one.
According to the concept of Mindset, which was developed by psychologist Carol Dweck, praise can be a powerful tool to foster growth and positive development in individuals. The key lies in delivering praise in a way that promotes a growth mindset rather than a fixed mindset. Here's a brief explanation of how to effectively give praise according to the Mindset approach:
Praise effort and the process: Instead of praising innate abilities or intelligence (e.g., "You're so smart" or "You're a natural at this"), focus on acknowledging and praising the effort the person put into their task or the strategies they employed to achieve their goal. For example, say, "I can see how hard you worked on this project," or "You did a great job in trying different approaches until you figured it out."
Emphasize progress and improvement: Recognize and celebrate the progress someone has made rather than just the end result. Encourage them to see challenges as opportunities for growth. For instance, say, "I've noticed how much you've improved in this area," or "You faced some tough obstacles, but you didn't give up, and that's commendable."
Be specific and genuine: Provide specific feedback about what the person did well. Generic or insincere praise may not be as effective. Highlight the particular actions or behaviors that stood out and express genuine admiration for their accomplishments.
Encourage a growth mindset: Reinforce the idea that abilities and intelligence can be developed through dedication and hard work. Encourage the individual to embrace challenges, learn from failures, and see mistakes as opportunities for learning and improvement.
Avoid comparisons: Refrain from comparing one individual to another, especially in terms of ability or intelligence. Each person's journey is unique, and comparisons can lead to feelings of inadequacy or superiority, fostering a fixed mindset.
Use "yet" to encourage persistence: If someone hasn't achieved a goal or mastered a skill yet, remind them that progress takes time and effort. Adding the word "yet" to statements, such as "You haven't grasped it yet, but I believe in your ability to get there with practice," can instill a sense of optimism and determination.
By providing praise that promotes a growth mindset, individuals are more likely to develop resilience, embrace challenges, and maintain a positive attitude towards learning and personal development. This, in turn, can lead to higher levels of achievement and a healthier approach to facing obstacles in various aspects of life.
The short version is simple: it is not about whether to praise or not—even the most careful teacher will still smile, give eye-contact, and give grades to high-performing students. Rather, the goal is to use praise with wisdom: to reward the right things, to speak precisely, to life people up and point them toward an even brighter future with each bit of praise.
I hope this gives you a sense for my position on carrots and sticks. We all use them. We might as well admit that.
And I’m not sure that’s a bad thing anyway.