Blocking effective treatment: Fraud posing as moral superiority, Part I.

“LOS ANGELES — A Los Angeles doctor was sentenced Friday to 14 years in federal prison for bilking patients out of more than $1 million by promising them that an herbal supplement she hawked could cure late-stage cancer and other diseases.”

A medical doctor convicted of selling snake oil. This really shouldn’t surprise anyone. The allure of positive reinforcement in the form of money and elevated status is highly corrosive to ethical behavior. One patient of Dr. Christine Daniel avoided treatment known to be effective for her lymphoma and died. Why? She relied on the promises, not of a fake, but of an academically trained scoundrel with real credentials. The same kind of behavior is common in the world of animal behavior and training. If you rely on the promises of an academically trained behaviorist or modern trainer using “scientific methods” and you are told to avoid treatment known by science to be effective, is it any different? Nope. In the law, it’s called fraud.

To fully appreciate this issue in the context of behavior and training it’s important to know a little about “learning theory.” First, it’s not really “theory.” It is a hodge-podge collection of unproven hypotheses and wild extrapolation. The foundational experiment examines the rate of response of a single behavior in one of two primative species. All else is projection. i.e. Because a rat will press a lever to get food, behavior analysts claim to know how to change the behavior of huge populations of humans. I’m not joking. The Assoc. for Behavior Analysis International is currently working with “green” activists to convince the world that global warming requires huge behavioral changes in the global population. Seriously. But they can’t train a dog to any professional standard. Hmmmm.


Tweek – the perpetually stressed kid on South Park. He has an eye-twitch, never can button his shirt and lives in perpetual fear – just like the rat in the chamber.

As bad as taking miniscule models and blowing them up to cosmic proportions is, it gets worse. Some of these hypotheses were established by monkeying with the experimental process to prove they were effective. That ain’t science. That’s a flimflam.

For instance, let’s say you need to generate numbers to get data to prove that your “science of behavior” is valid. (B.F. Skinner. Behavior of Organisms, 1938) You apply your experimental process to positive reinforcement and all is right with the world – connecting reinforcement increases behavior and you get lots of lovely lines and dots on a chart to show at a scientific conference. Then you decide to be objective and study the opposite of reinforcement – punishment. Oops. Punishment stops behavior from occurring. You’re chart is blank. You can’t show it on an overhead projector. Hmmmm. What now? Change the experiment. Reinforce the rat for pressing a lever. Then shock him silly. Now he’s sitting huddled in a four-wall box and that lever is looking him right between the eyes. He has to eat. His only way ot eat is to press the lever that just delivered a jolt instead of dinner. Aghhhhhh!!!! In terror and desperation, the rat presses the lever. Oh, my gosh! Food!!!!! The rat presses again. Food. Again. Food. Again. Food. Again. SHOCK!!!!!!! Repeat the sequence for a couple of weeks and voila, you’ve got numbers for your charts and graphs. You can now pronounce to the world that punishment doesn’t work because of “recent studies” that prove your flawed hypothesis. (B.F. Skinner: Science and Human Behavior, 1953)

Who cares that you weren’t really studying punishment as a discrete phenomenon? You were actually studying a model of abuse and calling it punishment. Who cares that you just distorted the experimental process in order to confirm your flawed hypothesis? If you pull this off, you get to be the 600 pound gorilla in the field of behavior analysis. Everyone will copy your experimental process because they, like you, will gain money and status. Yipee!!! A great, complex organization will grow and thrive that has all the appearances of science without all that logic and veritas – truth for truth’s sake. Thousands of behaviorists can get tenure, speak at scientific conferences and publish peer-reviewed papers, write books and stuff student’s heads with wonderful misinformation. No joke.

If you doubt my account of the model that is at the core of the behavior analytic conclusions about punishment, take a look at the abstract from a peer-reviewed paper from 1963 – Fixed Ratio Punishment by Azrin, Holz and Hake. This work was done in Skinner’s lab at Harvard under his oversight. He is cited in the paper as the author of similar, foundational experiments into punishment. I have underlined the important points.

Responses were maintained by a variable-interval schedule of food reinforcement. At the same time, punishment was delivered following every nth response (fixed-ratio punishment). The introduction of fixed-ratio punishment produced an initial phase during which the emission of responses was positively accelerated between punishments. Eventually, the degree of positive acceleration was reduced and a uniform but reduced rate of responding emerged. Large changes in the over-all level of responding were produced by the intensity of punishment, the value of the punishment ratio, and the level of food deprivation. The uniformity of response rate between punishments was invariant in spite of these changes in over-all rate and contrary to some plausible a priori theoretical considerations. Fixed-ratio punishment also produced phenomena previously observed under continuous punishment: warm-up effect and a compensatory increase. This type of intermittent punishment produced less rapid and less complete suppression than did continuous punishment.”

If you missed it, I’ll restate the obvious. They deprived the animals of food. (Put a prisoner in a dungeon cell and starve him for a couple of days.) Then they gave them food for doing a single behavior – the only behavior available. (Slide a tin plate with food partway under the door. The prisoner has to grab the plate with his hand to pull it the rest of the way. Every X times, the jailer bashes the prisoner’s hand with a truncheon and pulls the plate out of reach) In the pigeon experiments, they gave the food to maintain the responses because as we all know, punishment stops the animal from responding. That’s what Skinner said in his magnum opus. Isn’t that the logical reason to use punishment?

Even the last line of the abstract confirms reality – this kind of weird reinforcement and punishment combination wasn’t effective at suppressing a behavior. Consistent punishment is more effective at stopping behavior. Now we get to the meat of the matter. Azrin was using an experimental procedure that didn’t actually stop behavior. Why is that important? Because of what B.F. Skinner said in Science and Human Behavior, ten years earlier in 1953. (This also proves my claim that wild extrapolation is a part of behavior analysis. This project was done with six pigeons. That’s it. Of the 400 odd pages in of Skinner’s Science and Human Behavior, only 14 are devoted to punishment – all basically opposed to its use or study. At the 2009 Assoc. for Behavior Analysis International conference only ten out of 1500 presentations discussed punishment. In 2015 the ratio is 1500:0.) Here’s the seed of the fraud displayed for anyone with the ability to make logical connections. Here’s exactly what Skinner said in Science and Human Behavior.

“More recently, the suspicion has also arisen that punishment does not inf fact do what it is supposed to do. An immediate effect in reducing a tendency to behave is clear enough, but this may be misleading. The reduction in strength may not be permanent.” (Page 183. Free Press edition, 1965)

There are several parts of the fraud displayed here. First, the “recent suspicions” were based on Skinner’s flawed experimental procedure. A true parsing shows that he left out the word “false” – which should follow the word “recent.” Azrin’s last comment blows that out of the water – intermittent punishment produced less rapid and less complete suppression than continuous punishment. That means that any assessment of whether punishment works or not is dependent on how it is applied.

Second, why is permanent cessation of the behavior important? If analyzed by this criterion, positive reinforcement is an even bigger loser. It takes regular applications of positive reinforcement to maintain even simple pigeon pecking. Did Skinner know that? Of course. Behavior analysts routinely put an animal on an extinction schedule – they remove reinforcement to decrease behaving. Even the term is an attempt at distortion. Extinction schedules are a form of negative punishment. Why call them by a different name? So you can demonize punishment but go ahead and use one of its two forms. This is a fundamental error within behavior analysis. They invented the distinction between “positive and negative” punishment – but do not use their own terminology correctly when describing how behaviors form. When you use positive reinforcement to define an operant, a simultaneous decrease in other behaviors occurs from lack of reinforcement. There is no such logical thing as “shaping with positive reinforcement” without a concurrent negative punishment of all other behaviors. Want to check my statement? Let’s look at the definitions as accepted by behavior analysts.

Punishment Defined:
This definition of punishment is from Learning by A. Charles Catania – a pillar of behavior analytic thought. ( I loved his book. He’s got a great mind. It’s a fantastic starter for anyone interested in the science of behavior.)

“A stimulus is a positive punisher if it’s presentation reduces the likelihood of responses that produce it or a negative punisher if its removal reduces the likelihood of responses that terminate it.”

Also from Learning:
“Extinction: In operant behavior, discontinuing the reinforcement of a response (or the reduction in responding that follows this operation).

This is a distinction without a difference. What it means is that one cannot create behaviors with exclusively positive reinforcement. Negative punishment acts to suppress behaviors other than the developing operant. If you sit I give you treats. If you lie down you won’t get treats. Basically, the process includes both reward for doing the right thing and the removal of reward for doing the wrong thing. Why have a word, negative punishment, but not correctly identify it when it is displayed? The only answer is that the trainer doesn’t acknowledge the intent to punish the other behaviors. Now we are getting somewhere. The anti-punishment ideology attacks knowledge of or the intent to apply punishment. As long as we don’t call it by the “P” word we can ignore it. Once you mention punishment you have to shut up.

Going back to Skinner and the direct link to the title of this piece, why would obeying Skinner’s ideology cause the withholding of treatment known to be effective? Because behavior analysts know nothing of how to use punishment and actively work to block its use. In California, behavior analysts helped make it illegal to use punishment on anyone in state institutional care – regardless of context or the likelihood that the procedure would directly benefit the patient.

This enforcement of the behavior analytic orthodoxy means that people who have severe mental disorders such as self-abusive behavior are condemned to drugs and protective gear for life. That is despite the fact that punishment can stop the behaviors. I’m not making that up. Here is a recent post I got from a practitioner who heard me speak at an international behavior conference on the ethics and practice if including punishment in behavioral therapy. I offered and she accepted permission to post this as long as it was anonymous. Behavior analysts don’t like behavior analysts who leak the truth. In the most delicious of ironies, anyone who tries to reveal the reality of their ideology is instantly and severely punished. In the case of an applied behavior analyst that means no promotions, shunning and eventually a free pass to some other profession.

“Here in my current position, I often struggle with years and years of ‘reinforcement’ procedures that still have my students breaking their own noses and requiring 2-person holds – often on a daily basis, and an absolute resistance to even discussing the possibility of using a punishment procedure.”

Let that sink in a bit and I’ll hit you with Part II – animal applications.

NOTE: For those of you who wish to challenge my paper trail I would suggest that you are at a serious disadvantage. The paper trail exists as I have described it but it is only a tiny part of the literature that confirms what I say. If you want to argue, read this first – Ulrich, Wolfe & Dulaney, JEAB, 1969, Punishment of Shock Induced Aggression. Ulrich_jeabehav00146-0171 The scientifically validated, never contradicted way to stop aggression is with positive punishment. The specific modality is electric shock. Now tell me about shock collars and how non-punitive methods are scientific. Or, you can research topics like “Extinction Induced Aggression” where withholding rewards, the go-to solution of “positive” trainers and behaviorists,  triggered aggression. Question: If that happens, how does one stop the aggression created by the “positive” protocol? Just remember that the rule in the intro to my blog stands. If you wish to make a cogent comment you are welcome to join the discussion. If you wish to make a political statement, go elsewhere.





Leave a Reply

Your email address will not be published. Required fields are marked *