Author’s note. In 2009 I was pretty much fed up with scientists who distort reality. I have been a member of the Association for Behavior Analysis International since 1992 and have given many presentations and been an invited speaker more than once. I decided to give it one more chance and peel back the onion for those learned doctors. I submitted this handout to the review committee for the international conference. It was accepted. I gave the workshop. My demonstration assistants were two adult male Dobermans that had fought viciously several days before. They were still wearing a multitude of stitches from their battle. I had them lying side-by-side within ten minutes. They fell asleep about ten minutes later. Of the 1500 presentations that year, mine was the only one to demonstrate punishment and discuss it as a viable, humane, ethical means of modifying/removing unacceptable behavior. I also presented a paper on the topic and one of the only clinics in the world that uses punishment to stop serious behavior problems did several presentations about work. In all, that makes a comfortable 150:1 bias against any discussion of aversive control. That is why I get stuck talking about this topic. I hate deceit. Pretending that punishment is evil, harmful, risky, dangerous and traumatic leads to an unethical practice – withholding treatment and knowledge of treatment known to be effective. In that spirit I hope you enjoy this. (UPDATE: At the 2015 ABAI conference a catalog search pops up three abstracts containing the word “punishment”. None of them actually discuss anything substantive and one is actually an attempt to control behavior without punishment. That makes the odds up to about 1500:0. The proof of bias is plainly statistical in a group that is obsessed with statistics about behavior.)
Workshop: Theoretical and practical aspects of positive punishment and negative reinforcement.
Time: 6 hours
Description:
“If all you have is a hammer, everything looks like a nail.” – Bernard Beruch
In the field of behavior analysis, the hammer is positive reinforcement, albeit a velvet hammer. As token proof of this statement, consider this – at the 2008 CalABA conference, there were 143 presentations. 142 of them wielded the velvet hammer in one form or another. Only one presentation discussed positive punishment and negative reinforcement as valid, acceptable behavioral therapies and included practical advice on how to use them safely and effectively. I gave that presentation. I was swamped in the hotel lobby afterward by mental health practitioners who are not allowed to use, discuss or learn how to use punishment. While CalABA is merely the largest regional chapter of ABA, don’t imagine this hammer is a California fad. At the 2009 ABAI conference there will be 1500 presentations, posters, symposia, panel discussions, invited addresses and workshops. One workshop and one panel will discuss that actual process of punishment and how it might be used to modify the behavior or organisms.(Note: I wrote this before the conference. It occurred as I predicted with one small exception. The panel discussion was supposed to have three presenters. One guy didn’t show. So I got a grand total of 40 minutes to speak to the general conference attendees.) Rather than a conference on the unbiased examination and analysis of behavior, there is merely a collection of velvet hammers, attempting to hit everything, yet driving nothing. In a world that needs to drive nails, the hammer is conspicuously absent. The bias in favor of positive reinforcement permeates the discipline of behavior analysis at every level and across the globe and has since the mid-1930’s.
To quote the Frequently Asked Questions area of the Association for Behavior Analysis International (ABAI) website,
“Throughout his career, Skinner opposed the use of all forms of punishment; he advocated positive (sic) ways of changing behavior. “
A following statement further explains the position of behavior analysis regarding the use of positive punishment and by inference, negative reinforcement.
“Stronger forms of physical punishment, such as brief, mild electric shock are seldom used, and then only as a last resort with severe behavior disorders that have not responded to gentler procedures. “
Without arguing the use of the word “positive” to mean “gentler” and the implication that Skinner opposed the use of negative punishment – normally considered a gentle means of controlling behavior — these two statements point to the preference within behavior analysis for researching and developing “positive” methods and a complimentary bias against “negative” methods. To be precise, there is an obvious bias in favor of positive reinforcement and negative punishment (even if they don’t work) and an open disdain and denial of positive punishment and negative reinforcement. (even if they do work) This bias presents several serious problems.
The first problem is philosophical – how does a scientist reconcile a personal bias with the objectivity demanded by science? There are no other disciplines of science that pander to pet phenomena while implying that opposing phenomena are invalid. EG: Physicists do not prefer momentum while opposing inertia and biologists do not prefer mitosis while opposing meiosis. Instead, scientists examine nature objectively and strive for a balanced overview through specific, but complimentary, investigations. Selectively focusing on only half of a phenomenon (while discouraging research into the other half) is more common to theology than science. Behavior analysis has a track record of ignoring or demonizing less gentle behavioral tools without regard to objective analysis of readily available evidence.
Road Blocks to Objective Analysis:
Along with the philosophical conundrum of the “gentle” bias are two practical, logical road-blocks. First, the implication that punishment is only appropriate for severe behavior disorders and second, that positive punishment may only exist as a linear extension of a failed positive reinforcement protocol. Both of these concepts are at face value, illogical.
To deal with our first road block, consider this easily observed, well known example of positive punishment. The locking mechanism of an automobile seat-belt positively punishes rapid forward movement of the driver. No one would describe rapid forward movement to pick up a cup of coffee as a severe behavior disorder. No one would use brief, mild electric shock to stop this behavior nor would the shock be superior to the milder bump that occurs when the belt momentarily locks. Yet the procedure that causes the modification of the behavior is undeniably a positive punishment protocol. This procedure is not, by the definition of ABAI, a “positive” solution to the problem and would not be approved by B.F. Skinner, who opposed all forms of punishment and advocated positive solutions to behavior problems. The avowed philosophy of the founder of the discipline is ultimately gibberish when placed in a real world context.
The Road Block Revealed:
The road block is now obvious. A bias in favor of “positive” methods discounts this seat belt example as meaningless and not worthy of study – but using M&M’s to facilitate potty training is. One behavior prevents drivers from losing control as they bend over to prevent spilling a cup of coffee, while the other gradually reduces unwanted urine stains. One is potentially lethal, the other is merely embarrassing. The obvious conclusion is that positive punishment should not be and in reality is not restricted to severe behavior problems, is not always invasive and sometimes is the most logical, effective and safe first-resort to improving our quality of life. It also indicates that there is no practical use for positive reinforcement if you wish to stop a driver from grabbing a spilling coffee cup. The velvet hammer is an inappropriate and ineffective tool to solve this problem – and everyone knows it. That leads us to the second road block.
The “last resort” rule of punishment is even less logical than limiting punishment to the correction of serious behavior disorders. This dictum assumes that reinforcement and punishment are fundamentally similar in outcome and that the sequence of reinforcement followed by punishment is immutable. According to behavior analytic terminology, this cannot be correct. By definition, reinforcement (both polarities) increases behavior while punishment (again, both polarities) decreases behavior. These forces are logical opposites like up and down, left and right, accelerate and decelerate. To assume that positive reinforcement must always be used until it fails is like stepping on the gas pedal of a car in all cases – even if you see a stop light. Only when the accelerator fails to stop you, you may then hit the brake. i.e. Why would one assume that a behavioral solution must first embrace an acceleration of behavior? Is acceleration always the solution to every problem? Must we accelerate behavior until it fails before we attempt decreasing the behavior?
Once this reasoning is established, the road-block preventing objectivity is complete. We now have a behavior analytic community that assumes that “gentler” methods must be tried, even if they don’t work and there is no evidence that they have ever worked. There is no serious effort to study behavioral solutions that do not conform to the “reinforcement until it fails” protocol. Conversely, we are to withhold punishment protocols that have a history of effectiveness until the ineffective reinforcement methods have been exhausted. Are there no behaviors that plainly require a positive punishment protocol as the correct, initial treatment? Are all forms of punishment so risky, dangerous and unpredictable that we must never consider them? Is the death of the individual the only justification for using aversive control? Must aversive control automatically be intense, painful and dangerous in order to be effective? Must we only punish or reinforce, but never create concurrent contingencies – punishment for unacceptable behavior and reinforcement for a desired alternative?
The myriad of unanswered questions within the discussion of aversive control is astounding. All of these questions beg the more vital question and the purpose of this workshop – if punishment protocols are sometimes appropriate, who will apply them? Where can one learn to apply positive punishment to “serious behavior disorders” or even lesser behavioral issues? Does one learn to use this apparently dangerous tool by gradually shaping skills with less severe or less complex problems, which are not sanctioned by the “serious behavior disorder” philosophy, or does one only work on the most severe cases, sans experience?
In the world of behavior analysis, these questions are moot. Discussions of punishment as part of a valid behavioral treatment are scarce to the point of nonexistent. Many of the attendees of the 2008 CalABA conference were practitioners who use aversive control, secretly, every day. They have no resources to learn how to use it adroitly, with minimal risk to the client, because no one is teaching those things. The single presentation that discussed punishment was “standing room only” even though it did not offer continuing education credits. These people desperately need information about this process, yet are left in the lurch by the organization dedicated to compiling valid knowledge about behavioral control. Still, there is virtually no college course you can take to tell you how to punish self abusive behavior in humans or animals. You cannot major in behavior analysis with a minor in aversive studies. I know of only one university level class that actually has students practice negative reinforcement and punishment on living, complex organisms – and I teach it at Arizona State University in the Pre-Vet program, not the behavior analysis program.
While punishment is ignored and denigrated, humans and animals wait for the dissemination of practical advice and effective solutions to all levels of behavioral problems. If you desperately need to teach a puppy not to swallow socks – a common and invariably lethal, albeit mundane form of pica – where do you turn? Is the solution for this contained in Skinner’s Behavior of Organisms? Can you find it in Learning by Catania or Coercion and Its Fallout by Sidman? What text describes the establishing operation of making a neutral stimulus contiguous to a tangible punisher so that you can inhibit the pica in only one or two applications of positive punishment? What text describes how to inhibit the behavior when it occurs only in your absence? What source describes how to do this without adversely influencing the animal’s entire repertoire? Currently those discussions are not within the realm of academia.
The philosophical foundation of behavior analysis is supposed to be an objective, scientific perspective on the various phenomena that create and maintain behavior. It is not. Instead, with the ascendance of B.F. Skinner as the arbiter of “behaviorism” or “behavior analysis” a bias thwarted any efforts at objective observation or interpretation of behavioral principles. That bias has rendered modern education in behavior analysis ineffective at creating a true, balanced science of behavior. Punishment and negative reinforcement are discounted as topics of study and practical tools because of personal preference rather than scientific objectivity. The reality is that positive punishment and negative reinforcement are highly effective choices for solving many behavior problems, creating breakthroughs in learning and have little, if any risks to the organism – if you know how to do it. This workshop is intended for people who wish to develop a logical, objective view of behavior within the context of ethical and practical behavioral control of people and animals.
This workshop will acquaint attendees with the theoretical underpinning of positive punishment and negative reinforcement and practical applications of both. The workshop will include actual demonstrations of these behavioral effects on live subjects, both human and animal. ( Dear workshop review board member – did you metaphorically flinch when you read that last sentence? If so, you just experienced the logical justification for this workshop. If you assumed that a demonstration of this type would be somehow objectionable, dangerous, painful or unpleasant, you have simply affirmed my assertion of a pervasive, illogical and unscientific bias regarding this topic. In more than 20 years of offering behavioral therapy by veterinary referral, I have saved literally thousands of dogs from destruction using punishment and, less frequently, negative reinforcement in conjunction with positive reinforcement and negative punishment. I have never injured any of the animals I have worked with. ) Additional instruction will include video presentations of a wide variety of applications of positive punishment and negative reinforcement and a logical set of rules that govern the practical application of these two behavioral effects