Saturday, April 17, 2010

"Doing To" vs. "Working With" Alfie Kohn's Lecture on Threats and Bribes

Yesterday, we went to hear Alfie Kohn speak at the Montessori Centre Internationale (Denver, CO). It is always a pleasure to go to events like this, where you find yourself surrounded by like minded people and listening to a speaker you find intelligent and thought provoking. You might think that the influence of behaviorism and "discipline techniques" would be limited to parenting books and television reality shows, but attend even one continuing education lecture or early childhood education seminar not put on by Montessori professionals, and you will find that it continues to be the modus operandi of our profession- which isn't really that surprising, given how prevalent rewards and punishments are in society (merit based compensation, performance bonuses, public schools that reward everything from reading books over the summer to "paying" children with to sit quietly through class and allowing them to redeem their payment for prizes).

For example, a few weeks ago, I spent an entire weekend at a continuing education seminar essentially devoted to nothing other than different ways of setting up token economies (find out something a particular child really likes- say, stickers, M&M's, or in one case study they provided, a string of beads, then deprive the child of that and use it as an inducement for getting the desired behavior or stopping a "problem" behavior) for children with developmental differences. The seminar was put on by clinical social workers and "experts" in the treatment of this disorder, and it was really disappointing to find that they had nothing better to offer (as a side note, Montessori would have said we should observe the particular child closely to try to discover the reason for their behaviors and "follow the child"- which seems particularly insightful with this disorder, given that the children often have differences in their emotional regulation and vestibular processing that causes them to have "bizarre" behaviors like spinning, chewing incessantly, etc- behaviors that are not problems, except perhaps for us, but help them with self-regulation). I guess her insights into developmental differences are radical even by today's standards! Anyhow, I digress.... back to Alfie Kohn.

One of the things that I really enjoy about Alfie Kohn is that all of his books are supported by a vast body of scientific literature (reading his footnotes and endnotes is as exciting as reading his books); additionally, his approach has rather obvious affinities with the Montessori method (respecting the liberty of the child, her aversion to both punishments and rewards, giving the child freedom of choice, engaging the child with purposeful activity, and the importance of an aware adult). Alfie Kohn also explicitly gave permission for notes from his lecture to be disseminated and shared. So, in that spirit, I am sharing them here with you as a little thank you to our parents who were inconvenienced by having to find other childcare arrangements so that we could continue our education. I hope that it inspires you to pick up one of his excellent books (most of the information contained in this lecture is covered in depth in his book, Punished By Rewards), check out one of his articles on his website (all of his articles can be accessed from the site for free), or re-examine how we (as parents or teachers) respond to children and how we can reach out to them and interact with them in a more compassionate, authentic, humane manner.

Alfie Kohn is speaking again today at Naropa University in Boulder (I believe tickets are still available, although the topic is "The Case Against Homework," something less germane to our age group).

For more information, please visit his website at

Notes From Alfie Kohn's Lecture: Beyond Threats and Bribes
4/16/2010 at Montessori Centre Internationale (Denver, CO)
(I have tried to reproduce these with fidelity to his lecture; however, it can be hard to separate what you hear from your reactions to it. Phrases or thoughts may include my own embellishments; errors, misspellings, and omissions are my own).

I. Introduction

It is a well established fact that learning is not merely absorbing information, but the process of actively constructing meaning around information (Piaget, Vygotsky, Montessori); as a result, the best teachers are the best listeners.

There is a metadiscourse to his presentations; A.K. likes to begin his lectures by allowing participants to construct their own meanings (not merely absorb facts and opinions). He begins by presenting a study in which the researcher got unexpected results and giving the audience time to formulate a hypothesis which would explain them.

A.K. cited two studies, one from Arizona State University and one from the University of Toronto, which concluded that children whose parents offered positive reinforcement regularly (whether the reinforcement came in the form of stickers or verbal reinforcement) were less caring, less sharing, and less helpful than their peers. The negative effects on helpfulness were most pronounced when the positive reinforcement was directed at sharing (when they were rewarded for sharing). These children were the most selfish and least likely to help others.

When we speak about "goodness" in adults, we are usually talking about people who are ethical, honorable, or altruistic. By contrary, when we speak about "goodness" in children, we mean compliance/obedience (children who do what they are told). Over the past few years, the "techniques" to get obedience have shifted from punishments, especially corporal control (spanking), to verbal inducements (the ubiquitous "good job"). However, few people have considered that these approaches often backfire and encourage defiance or excessive compliance and capitulation.

To take one example, Illinois researcher Leanne Birch has done several studies on food and control (parents who force children to finish their plate, eat only at specific times, eat this before that, etc). Her research has discovered that children have a good natural gyroscope (homeostasis) of nutritional regulation except when over-controlling parents fetishize/ interfere (which causes them to be more likely to be overweight). This doesn't mean that children always eat what is best for them, but in homes where nutritious (not excessively processed, sweet, salty, or fat -laden) food is offered regularly , children may eat less at one or meals, but over time they tend to self-regulate to get the calories and nutrients their body needs.
What is true for food, is true for other things. Additionally, how do children learn to be moral if people are always telling them what to do?

II. On Punishment
There are two fundamental techniques for control. The first is punishment. Is punishment effective? Research shows that if the punishment is severe enough and the probability of it being used is high enough, then punishment can be effective in getting temporary compliance.

Kohn likes to begin by looking at punishment logically and asking these questions:
1) By making children suffer in some way, by making them unhappy, how likely is it that we will make them better people?
2) If punishment is a good idea, why do we have to keep using the same punishments (sometimes, even for the same infraction), over and over on the same kids?
3) What are the alternatives to "doing to" (his word for punishments)?

Kohn also likes to remind people that the euphemisms we use for punishments, calling them "consequences" or "logical consequences" does not change the fundamental nature of them. They just make us feel better.

The research is very clear. Behaviorism only works for temporary compliance. Additionally, studies show that punishment emerges as the cause of many universal behavior problems. In fact, punishment is a more unique variable contributing to problem behaviors than all other factors. So it is somewhat ironic that we know many behavior problems are caused by punishments at home, and yet we respond to these same children in the schools with more punishments.

The research is strongest against corporal punishments. Children who are recipients of corporal punishments are more likely to be aggressive, depressed, and to grow up to repeat the cycle (using corporal punishments on their own children). Punishment teaches children that when they have more power, and can hurt others, they can do what they want (voluntarism- the strong do what they can; the weak do what they must). It is destructive to relationships as these children often harbor resentment and feel dis-trusted and un-loved.

In fact, Hallis Miller, argues that we don't do it (punishments, especially corporal) to children in spite of the fact that we had it done to us, but because it was done to us. On an unconscious level, we desperately need to believe that our parents hurt us/did that to us out of love or concern for us. We do it to our children to make it so and to fill our own psychological need.

What is "time out?" It is forcibly isolating a child from their peers. Exiling them when they need to feel part of a group or social community. It stigmatizes and humiliates them in front of their peers. Many early childhood professionals have issued a mea culpa for inflicting this on children when they reflect upon how they want to be treated if they make a mistake in judgment. A self-imposed time out is good for adults who are out of control and might be a good way to teach children a positive coping strategy- but only if the child decides where to go during their time out, what to do, and when to come back.

Unconvinced? Consider the etymology of this euphemism. The term "time out" hearkens back to BF Skinner and refers to a time out from positive reinforcers used on lab animals (for example, when an animal isn't doing what we want in spit of the administration of positive reinforcers, we plunge them into darkness for a period of time).

Why does punishment prove to be so destructive and ineffective beyond temporary compliance?
1) It makes the recipient mad. It fills them with rage, defiance, and a desire to get even. It is not a coincidence that victims often go on to become victimizers.
2) All that it teaches is a lesson in power distribution. It distorts the intended lesson (be kind to your friends, etc) and instead what the child learns is how to use power to make others capitulate.
3) Eventually it loses its effectiveness. Consider a teenager who cannot be physically controlled.
4) By staking everything on power (a foolish strategy), you lose your opportunity to develop a relationship based on influence.
AK quotes Thomas Goodman (developer of Parent Effectiveness Training): "The more you use power to control your kids, the less you influence them."
We want to be a child's caring ally, to offer them guidance and support, not an enforcer. Punishments erode a child's sense of security and trust. They encourage children to lie and keep their distance (not to solicit help and advice when they need it).
5) They distract children from dealing with important issues. Rather than thinking about the problem to be solved, but as something that will be punished. Punishments cause children to think about avoiding detection, lying.
6) All "consequences" (even "logical/natural" ones) make children more self- centered. Where we should be teaching children to consider the consequences of their actions on other people (we don't hurt people because we don't want to hurt them), punishments cause children to think about their own self- interest (what will happen to me).

So, if punishments are so destructive, why do we use them at all? They are perceived to be efficient (particularly in group education settings- we believe that working through the issue would take too much time so we banish the offender instead), they may serve as a release valve for out own tension/serve our own psychological needs, we don't know what else to do, they are easy, we are afraid of what will happen if we don't, and because our society, especially parenting advice, is bound by dichotomous thinking (ignore or punish, be permissive or punitive, etc). Whereas "working with" (A.K.'s alternative to punishment) takes time and talent; punishments are easy and formulaic (often, they can be reduced to repeatable "techniques").

A.K. then cited a study that showed that most 2 year olds will repeat an unwanted behavior regardless of the intervention used (whether it was a smack on the bottom or a conversation) hours later. However, there is a double standard here. When the parent tried to "work with" the child and it didn't work to change future behavior, we often say that it is because the parent was too lenient and didn't punish the child (we blame the parenting approach); on the other hand, when the child was punished and it didn't work, we say that is a reflection on "how bad" the child is (we blame the child).

So-called "natural consequences" can be used in a way that makes sense (a child stays up too late so they are tired the next day or doesn't clean their room and cannot find their belongings), but most of the time, parents use them as a euphemism for punishment (for example, you are running late so now I won't give you a ride); in most cases, it boils down to mom/dad could have helped me but didn't.

III. Rewards

Do rewards work? If the reward is great enough and the probability of getting it is great enough rewards are effective, but only at a significant cost.

Two dozen cross-cultural studies prove that if people are given a task and one group is offered a reward for completing it, the group with no reward will do a better job. This was well-established as early as 1971 when Janet Spence of the University of Texas (now head of the American Psychological Association) concluded that "rewards have effects that interfere with performance for reasons we are only beginning to understand." No long term improvement in performance has ever been discovered from use of rewards! Research shows that when rewards are offered (including grades and merit based pay), people remember facts for a shorter period of time, are less creative, lose their intrinsic interest in the thing being rewarded, are less likely to do the thing being rewarded on their own (without the reward), and the quality of work is more superficial.

At best, incentivizing does not work; however, often is is worse than doing nothing. Bribes, like threats, are counter-productive. Why are rewards so harmful?
1) Rewards are sugar-coated control. Rewards are not the opposite of punishments, they are two sides of the same coin.
2) Rewards create a focus on doing things for your own self- interest.
3) Rewards can generate feelings of inadequacy when you don't get them (if all that is important is getting an 'A,' but you can't, why strive for a 'B'?)
4) Rewards create a glass ceiling. There is no incentive to strive beyond the reward (if you have an 'A' in the class, why pursue your study farther?)
5) Rewards teach risk avoidance (why push yourself to read a really challenging text or write a challenging paper, if you can get the 'A' by doing something easier, why bring an important problem to your employer's attention if your bonus is focused on the appearance of things going well).
6) Rewards reduce the likelihood of playing with possibilities that might pay off (if the mouse knows one path to get to the brie, why look for a shortcut), reducing creativity.
7) Rewards pervert relationships. Rather than seeing your teachers or bosses as a caring ally, you begin to see them as reward dispensers. They create a culture of 'pleasers.'
8) Rewards depend upon surveillance (being seen doing the desired behavior); as a result, they encourage people to be superficial (it is not important to be good, but to appear to be good). They teach the form of morality, without it's substance.

We discussed how rewards hurt vertical relationships; now, let's discuss horizontal relationships. What's worse than a reward? An award. Making a reward artificially scarce so that children actively hope for the failure of their peers.

Little kids are so desperate for our approval, it is abhorrent to use it for our own convenience.

Here is Kohn's Law: The number of times you see or hear the words behavior in a resource (whether a teaching or a parenting resource), the less useful the resource is.

Rewards don't work because they ignore reasons. The problem with behaviorism is not the use of reinforcers, per se, but the belief that we are nothing but behavior. Behaviorism is formulaic; it doesn't ask why. It portends to be able to address perceived problems without knowing the underlying causes (or even caring about them). For example, a behaviorist doesn't need to know why a child isn't falling asleep, they just put the child to bed and slowly lengthen the amount of time between the child's cries and the adult's response or they give the child a new toy if they sleep without crying every night for a week. But shouldn't our response be different depending upon the reason? Is the child hungry? overtired? being put to bed too early? afraid? too busy processing something from their day? lonely? wanting more time with mom/dad? etc. To "work with' on the other hand, we need to know why, we need to understand the child's needs and values. A.K. is interested in the child- in how things look from the child's perspective.

People tend not to do as well when they are trying to get a reward because their motivation has changed; the extrinsic motivators undermine the intrinsic motivation (which is more powerful). Additionally, research shows that the more you reward people for doing something, the more they lose interest in doing it for it's own merits. For example, kids who are trying to get good grades are intrinsically less interested in what they are learning. A.K. cited a study in which children were rewarded for drinking kefir. The children in groups that got verbal rewards (the ubiquitous "good job") and tangible rewards drank more kefir during the study than the control group, but reported liking kefir significantly less than the control group when they study was finished. Similar results in other studies show that rewards make kids less likely to do the thing being rewarded on their own. A.K. jokes that he doesn't care if kids like kefir, but substitute reading or math as the thing being rewarded and the problem becomes immediately apparent.

The research is clear: the more you want you kids to do something, the more you should avoid rewarding them for doing it.

A.K. says he is not saying that praise is morally equivalent to child abuse, but it is important to acknowledge that they are on the same continuum.

IV. Alternatives to Threats and Bribes- "Working With" vs. "Doing To"

The starting point for "working with" is to think of your goal; it depends upon our own beliefs about the child and about human nature ("Begin with the End in Mind"). A.K. wrote a book called The Brighter Side of Human Nature: Generosity about teaching children generosity.

There is an empirical correlation between people with cynical beliefs about human nature and greater use of punnishments and rewards. People who believe that when kids do something nice it is a fluke, tend to feel the need to swoop in with an artificial reward; whereas people with a balanced or positive view of human nature, merely feel they need to support children and their natural generosity.

Nell Noddings (Stanford University) says: "We should attribute to children the best possible motivations consistent with the facts."

If a young child is not behaving the way we think they should it is possible they do not have the memory, skills, etc to do what we are asking.

How Can We "Work With?"
1- Attend to the relationships among kids; foster a caring community
2- Observe and identify things that get in the way or peaceful relations.
3- Eliminate competition- see for lovely games that involve figuring things out together; Terry Orlach from the University of Ottawa has a great book Cooperative Sports and Games. Avoid teaching children that other people are obstacles to their success.
4- Think in the plural ("we" "our community"); be inclusive and don't isolate or alienate.
5- Find engaging stuff to do! The curriculum matters! If it isn't interesting and engaging to them, children get into trouble. When schools become standardized test prep centers, children tune out, burn out, act out, and drop out.
6- When kids don't do what you want, stop and ask yourself: What am I asking them to do? Why- for what purpose? Is it developmentally appropriate? Are they engaged or bored? What do I want them to learn from doing it? What is the consequence of them not doing it? Is it more important than my relatioship with this child? Does the child know that I care?
7- Respond to problems in a way that reflects your care and concern for that child ("I wouldn't let anyone in here do that to you, so I can't let you do that.")
8-Talk less- ask more; ask questions and avoid judgments. Seek first to understand then to be understood.
9- There has to be an unconditionality. No matter what they do, children need to know that you will never, ever, stop caring about them. Your care and concern do not have to be earned.
10- Give them choice! Children learn to make good decisions by making decisions, not by marinating them in praise.
11- Don't be afraid to give up control and attention to the children- abdicate control gracefully. Involving them in the process of deciding what to do or how to solve problems supports their social, emotional, anc cognitive development.
12- A colleague once told A.K. "My job is to be as democratic as I can stand." Push yourself to stand more every year.
13- Don't steal their own pleasure with their accomplishments by praising or substituting your own judgment.
14- If someone does something nice, rather than praising them, consider directing their attention to how their actions made the recipient feel ("Wow, it looks like ___ is smiling; she really likes that sandwich you shared with her.")


  1. I always find it disconcerting that a child can be in school for six hours, functioning beautifully without praise, rewards, or punishments, and yet as soon as they get in the car with their mom, the adult says: "I have a treat for you at home because you were a good girl at school." Now, is that truly necessary? Ugh. Thanks for the wonderful and thorough notes, I am a huge admirer of Alfie Kohn and wish more parents would pay attention to his insights.

  2. Thank you! I totally agree as a Montessorian!