Teaching Inhibitions: Safe, effective punishment revealed.
Note: In recognition of the mainstream philosophy of modern dog training I would like to include this preface. This article covers techniques and principles that can prevent a dog from biting your leg as you run an agility course, attacking another dog or simply skipping out of the ring during a run. If you perceive a bias in favor of punishment over positive reinforcement, it is the result of the specific nature of the topic rather than any personal preference of mine. I use tools that are most likely to benefit the dog. I don’t preselect what I will and won’t use, based on my sensitivities or some ideological devotion to long-dead behaviorists. Behaviorists whose learning hypothoses have never produced a dependable working dog. The dog comes first. The client comes second. Skinner, comes last unless I am training rats or pigeons in a tiny little box. The problem is that for more than 20 years, the growing drum beat of “all positive training” has stifled discussion about teaching inhibitions. The information contained here isn’t speculation, it’s the result of 25 years of practical experience with literally thousands of dogs. During that time I have never injured any dog and always left them with their tails wagging. As you will read, as in this case study, I have enhanced the lives of dogs by using methods that directly benefit them without possibility of harm.
About 15 years ago I spoke at a training conference in Chicago. One of the people in attendance was Dee Bramble. Dee and her sister, Jill, own “Over Rover” – in my opinion, one of the best agility schools in the country. Dee’s problem was simple – her Cattle Dog, Boomer, routinely bit her legs during agility runs. If you’ve never worked with a Heeler you need to know two things to understand this problem. First, cattle dogs are bred to drive other animals by biting their heels – hence the term heeler. Boomer was a heeler in every sense of the word. Second, when cattle dogs bite, they bite -- it’s not a nip. If a 50 pound heeler nails you in the calf, you are going to be inspecting the tundra from close range.
Dee’s immediate problem was that unless she could stop Boomer’s aggression, she would have to retire him at an early age. She had consulted a good number of trainers with no success. I fixed the immediate problem in about 20 minutes, with a 10 repetitions of my solution for aggression. Dee followed up the process with some touch-ups over the next few weeks. End of problem. Boomer went back to competition but didn’t bite Dee in the ring.
A couple years later, I presented an agility camp at Over Rover. One of the dogs was an adolescent Malinois who had the habit of darting out of the course and attacking other dogs. After “fixing” him in a short, 15 minute session, we sent him around the course and through a tunnel. As will ever happen when you think everything is under control, Boomer had entered the course unseen – he lives there and sometimes thinks he owns the place. Boomer entered the tunnel at exactly the same time as the Malinois, from the other end. These two male dogs with a history of aggression had a head-on collision inside the close confines of the tunnel. No fight occurred. Both dogs turned 180° and came out the way they went in. This is the true nature of an inhibition – in the presence of all the factors that would normally trigger a behavior, it doesn’t happen.
In case you’re wondering, I didn’t use any magic formula or dominance/pack-animal stuff to figure out what was going on -- I certainly didn’t whisper at either of these dogs. In each case, I picked the moment when the unacceptable behavior started and said the word No -- and then threw a “bonker”, a rolled up towel, so that it bonked them on the head, from about 10 feet away. i.e. I correctly, precisely, safely and humanely punished their aggression. To imagine what it’s like to be bonked, consider what if feels like to be in a serious pillow fight. I use a rolled up towel for three reasons; it poses no risk of damage to the dog, dogs instinctively hate projectiles and they can be thrown precisely. More than anything you are startling the heck out of the dog, which interrupts the behavior and causes an inhibition to form. I continued to apply this consequence each time they displayed the behavior, until the behavior stopped completely. In each case, it took about four bonks. After the behavior was inhibited, I used lavish praise and treats to get their correct behaviors up and running again. (Please read that last sentence again before you proceed. The follow-up reinforcement for correct behavior is a huge part of the long-term solution to aggression) In each case, the immediate effects of being bonked (caution, fearfulness and confusion) were transitory and short lived. These emotional reactions are normal, acceptable and completely harmless. More importantly, the unpleasant event of being bonked stopped the aggression and allowed me to guide the dog to a new course of action. As you use positive reinforcement to maintain the new inhibitions, these emotional reactions are quickly forgotten and entirely inconsequential. Neither dog was done with the training until they were confidently back doing their work, minus the objectionable behavior.
Contrary to what you might think, this process didn’t cause any of the dire predictions that are so common in modern dog training circles. The “bonkings” didn’t slow their agility runs. The dogs didn’t attack me or anyone else when they got bonked. They didn’t become hand-shy. They didn’t receive any wounds caused by their uninhibited fighting. The reality was that there were no bad side effects, merely beneficial, immediate and direct effects. Boomer stopped attacking Dee when they ran agility and the Malinois stopped exiting the ring to attack other dogs. The price tag for these improvements was small and nothing like the damage their aggression could cause.
The Broad Picture:
Obviously, these dogs’ uninhibited and unacceptable behaviors aren’t unique. In fact, these behaviors are perfectly normal and all too common, even among serious trainers. Contrary to the popular concept that “violence begets violence,” in reality, it is DNA that begets violence. All dogs are born with the potential for aggression. Many people simply don’t imagine their dog will ever do anything like that. That explains the famous last words of virtually every dog bite or disobedience -- “Gee, he’s never done that before.” In most cases, this statement is entirely true – but not particularly savvy. Anyone who has been around dogs should assume that unless you specifically inhibit undesired behaviors, they may lie dormant for the dog’s lifetime. Expecting a dog to magically inhibit his own natural behaviors is wishful thinking, but it has no basis in reality. Instinctive behaviors remain dormant until the correct combination of events makes the behavior functional -, then it happens. This is the same whether the behavior is aggression, scavenging or chasing a cat. (Note: By “specifically inhibiting”, I mean connecting an unpleasant event to a behavior that prevents it from happening in the future. The scientific term for that process is positive punishment – an effect that is popularly portrayed as dangerous, traumatic and evil, all rolled into one. Ironically, scientists do not study this topic and are not really experts on the topic of the practical use of aversive control. The reality that most aggressive dogs will die if their behavior doesn’t stop and that “positive” methods can’t stop serious aggression seems to be lost on modern dog training philosophy. By definition, positive reinforcement cannot stop any behavior. )
A common occurrence at dog events and dog parks is an unexpected fight between seemingly obedient dogs. In some cases, one of the dogs has a history of aggression and the other is a sweet, innocent, completely passive creature who is suddenly forced to fight for his life. As a result of the incident, the recently, never-aggressive dog often becomes wary of other dogs and starts to initiate aggression before he can be attacked. His primary motivation for future violence is fearful self-defense. This creates an odd irony. His owner is unlikely to use any form of correction to stop this new aggression because it might scare the dog – a dog who is already terrified far worse than any punishment protocol can evoke. In essence, the dog’s emotions have suddenly been elevated to a status that ruins any chance of fixing the problem. This would be like refusing necessary medical treatment because it might scare your dog. Logically, if pain and fear were always traumatic, all dogs would be terrified of the vet. Equally logical is the notion that no one assumes that withholding the necessary discomfort at a vet clinic is a justification for withholding treatment. I would suggest that since most canine aggression is fatal, it deserves the same logic applied to its solution. Fear and pain is not automatically an ethical reason to withhold treatment for either medical or behavior problems.
If I were to limit my methods to positive reinforcement, I would run into a very big snag. The now-aggressive dog’s fearfulness makes the use of positive reinforcement impractical to pointless. A scared dog won’t take treats and, at best, is annoyed by your attempts to pet and stroke him. Some dogs will snap reflexively at any attempt to physically restrain them when they are near the threshold that triggers their attack. A logical analysis would yield this result -- you have a dog who is scared silly, won’t take treats and is focused on killing any dog who steps over his particular boundaries. In most cases, people do nothing and hope for the best or stop taking the dog to dog events.
If you read dog books, magazines and internet sites you will probably assume that I’ve forgotten the two most common ways to solve behavior problems -- “gradual desensitization”, getting a dog used to being attacked by giving him treats, or “counter conditioning” -- teaching the dog an incompatible behavior that somehow prevents the dog from defending himself. These two methods are rarely adequate to stop the behavior. As I said, aggression, along with foraging, eating inedible objects and a host of other unacceptable behaviors are the result of DNA, not learning. These are normally occurring behaviors. The reality is that positive reinforcement cannot inhibit normally occurring behaviors. To make this point clear here’s a logical parallel to the “treats to prevent aggression” perspective.
Example: Just as with aggression, leg-lifting to urinate is a genetic behavior triggered by age and environmental cues, such as the smell of another dog’s urine. To use gradual desensitization, one would use dilutions of male dog urine to offer a level of smell that doesn’t trigger leg lifting. If the dog doesn’t lift his leg in the presence of the weaker smell, you give treats and affection. Gradually, you increase the potency of the urine until you are at 100% urine with no leg-lifting. Voila, instinctive leg-lifting vanishes! Not.
If you select counter-conditioning to stop normally occurring leg-lifting, it rapidly becomes laughable. Does anyone imagine you can stop a mature male dog from lifting his leg by giving him treats for an incompatible behavior like “roll over?” Even if you could possibly slow down the leg-lifting this way, counter-conditioning doesn’t remove the dog’s ability to instantly regenerate the behavior at a later date. At best, this solution requires perpetual vigilance, a constantly hungry dog and a tolerance for periodic recurrences of the behavior. This is an annoyance if the behavior is leg-lifting and disastrous if the behavior is aggression.
Needless Embarassment vs. Artful Pride:
Aggression in the ring, like Boomer’s, is only one of several embarrassing problems you’ll see in the modern training world. A few years ago a top agility competitor left the ring dripping blood from her arm – her Border Collie was so jacked-up at the end of the run that he leapt up and bit her forearm. Her solution was to wear long sleeves for the summer. Don’t think this is limited to the sport of agility. Several years ago at a major competition of police dog handlers, the same thing happened. A very jacked-up Malinois was so frustrated at being told not to bite a bite-suited cop that he leaped up in the air and attached himself to his handler.
On a lesser scale of importance, but equally common, you will also see dogs exit the ring in the middle of a run or go to sniff a treat that was dropped on the ground while their handler screams commands that aren’t likely to be obeyed. Sometimes it’s to make a beeline for another dog. Other times it’s merely to frolic and do whatever is on the dog’s mind. In an all-positive world, these time-wasting and frustrating behaviors are a fact of life for a significant number of people. That’s because positive reinforcement cannot inhibit normally occurring behavior. The only behavioral effect that can stop an existing, normally occurring behavior is positive punishment. Realistically, you can learn to use this tool to safely teach your dog to work to its full potential – or not.
You may train with people who believe that any use of punishment will ruin a dog. I would suggest that most of these people are basing their philosophy on good intentions and hearsay. I would also suggest that they likely have no actual experience at using punishment, correctly, safely and effectively. The use of punishment is almost universally assaulted and equally unknown. To my knowledge, there has only been one college level course that included a discussion of why, when and how to use punishment - and I taught it. Likewise, I am the only person to ever present a hands-on workshop at a major scientific conference. I stopped two adult Dobermans from trying to kill each other. There were 1500 presentations at that conference - and only one that demonstrated the real rules of using aversive control.
The rules that I will outline here cover how to stop behaviors without any risk to the animal – meaning how to punish correctly. I will assert that attempting to punish by throwing all manner of nastiness at a dog is merely abuse and has nothing to do with this topic. The ancient Romans recognized the illogic of the current rhetoric opposing all forms of punishment. “Abusum non tollit usus” – the abuse of a tool is not an argument against its proper use. If you are interested understanding this issue fully, read on.
The Proper Usage:
The first thing to understand about inhibiting behaviors is this: positive reinforcement (or withholding positive reinforcement) cannot create inhibitions.
To show you how important I think this is I’m going to repeat myself so you can’t possibly miss the point. Ready? OK. Again.
Positive reinforcement (or withholding positive reinforcement) cannot create inhibitions. It doesn’t matter how many letters you put after your name, where you were trained or what species you work with, this is a biological fact and inescapable. If it seems that I’ve been a bit forceful in making this point, I assure you it’s a critical link to solving the overall puzzle of behavioral control and it is rarely given the emphasis it deserves. To put it further in context, let’s take a step back and look more closely at the underlying principle, inhibition.
OK, what exactly is an inhibition? Behavior analysts and dog behavior experts rarely use the word but it’s a pretty apt term for our purposes and one that virtually everyone understands.
Here’s Webster’s take on it…
Inhibition: An inner impediment to free activity, expression, or functioning: as a mental process imposing restraint upon behavior or another mental process (as a desire)
To paraphrase Webster, an inhibition is a road-block that stops a behavior from happening. I would add that to be at all useful, an inhibition must include this criterion: even in the presence of all the things that would normally trigger the behavior, the behavior doesn’t happen. If you listen to most dog trainers and the majority of dog owners, they invariably make excuses about their dog’s behavior. “He does a great stay, unless he hears the door bell.” Or, “She was attacked by a Bearded Collie once. Ever since then, she’s been really bad around shaggy dogs.” If you wish to stop making excuses for your dog’s bad behavior you must learn how to stop it in the real world – not a world devoid of shaggy dogs and starting lines.
The most important thing to remember is this -- positive reinforcement strengthens behaviors through pleasant means, it doesn’t stop them. In most cases, it is impractical, if not impossible to remove all the positively reinforcing influences that perpetuate a behavior. Even if you succeed temporarily, there is no inhibition that prevents the behavior from instantly regenerating when reinforcement again becomes available. That’s why a single guest in your house who boogie-woogie-kissy-kisses your dog will destroy your attempts to withhold reinforcement for jumping on guests. In the broader world, consider how easy it is for one puff of tobacco smoke to ruin someone’s attempt to quit smoking. True story, ask me how. If you doubt that behaviors don’t go away merely as the result of not reinforcing them, try to remember any of your elementary school teachers, or your house when you were a child. I doubt that anyone has rewarded you for coming up with that information in a long, long time. Did you forget it after all these years? Nope. Hey, I thought if you stop reinforcing a behavior it will go away? See how silly that sounds when you plug it in to the real world?
This brings us to the crux of the issue. Understanding how to inhibit behaviors spans everything from allowing a dog to realize its potential in the agility ring to potentially saving its life. I’m not exaggerating about that last part. I spent eight years euthanizing dogs who exhibited no worse “problem” than heeling the children around the swimming pool or jumping up when greeting people. Dogs who can’t adapt to humans die at a rate of nine to one to the tune of tens of millions, annually. You may decry the triviality of why people get rid of dogs, but you can’t beat the horrific odds by lamenting a tragedy. The only way to solve these problems is to learn to stop the behaviors that make dogs unwanted and impossible for normal people to live with. If you think that people who participate in dog sports aren’t involved in this process, think again.
When a “newbie” at your agility club ends up with a dog like Boomer, what are you going to do about it? Are you going to suggest that the dog be put in a “time-out” every time the behavior occurs? Remember, the behavior isn’t being driven by reinforcement and won’t stop if you somehow remove reinforcement. i.e. You can’t make a heeler stop heeling by putting him in a crate for five minutes. Even if you could, you’d be wasting the better part of your time with your dog in a crate rather than shooting for a clean run. Don’t assume that I’m only talking about in-ring aggression. A newbie whose dog exits the ring constantly or stands around sniffing discarded treats is likely to get a different dog or simply stop doing agility out of frustration or embarrassment. When the club’s “Never say No!” policy fails to stop the behavior, what happens to the dog? Assuming they are going to keep a dog that can’t participate in their new hobby may be wishful thinking. If you can stop the unwanted behavior you have a win-win situation. If you can’t, the newbie is on her own and may decide to take up quilting.
It takes moral courage to apply an unpleasant consequence to a dog’s behavior. It’s easy to be a coward, ignore the behavior and learn to live with botched runs, bitten children and “dog at large” citations. The choice is yours. The logical result of programs that use nothing but pleasant consequences is simple. To express it best I must use a hackneyed cliché that is perfectly appropriate for this discussion. Failing to use unpleasant consequences to stop behaviors such as jumping and biting leads to a consistent result -- you literally kill the dog with kindness.
This article is not meant to use as an outline to effectively inhibit behavior. It is meant as a reminder for you that the process requires learned skills and the practice time to become proficient at those skills. It is also not meant for someone to skim the information and then run around “bonking” dogs – and that’s why it ends here. For more information regarding inhibitions you can go to my website at www.clickandtreat.com or contact me directly at firstname.lastname@example.org
Last, but not least, while working in shelters, I personally killed thousands of dogs whose only fault was that they jumped up on a prospective adopter or ripped down some drapes after they found a good, albeit momentary, home. When I decided to learn about fixing their behavioral problems, I held to the same ethics as the vet who was my mentor for three years – Dr. H. Dean Bauman of Corvallis, Oregon. The ethic wasn’t “do no discomfort” or “cause no suspicion or fear.” The ethic was, first, do no harm. The second ethic is from Larry the Cable Guy – “get-r-done.” Using these guidelines I have saved thousands of dogs over the last 20 years in an effort to balance the thousands I couldn’t save when I worked in shelters. About 95% of what I do is to apply positive reinforcement to create new, functional and acceptable behaviors. About 3% is kissing puppies and rubbing dog bellies, two of my favorite pastimes. The final 2% requires that I do, not what I love to do, but what the dog needs to stay alive.