Why I Don’t Believe In Clicker Training
Call Today!
Sam Basso
PHOENIX , AZ AREA: (602) 708-4531
OR, if you are out of this area, inquire about a telephone or e-Lesson
Email: [email protected]
“Clicker training fanatics make the same mistakes as all dedicated reductionists: always pretending to know more and more, and devoted to knowing about less and less. They are more concerned about their method than the whole dog, and thus miss the entire point of why a dog does what it does, and what dog training is all about.” – Sam Basso, 2013 [Please Read: What Is Operant Conditioning?]
Let’s start with the fact that I DO use clickers training some dogs. Yet, clicker training is not the entire solution for training dogs from start to finish, and they aren’ the right tool for every training situation. This article is to explain situations where I don’t believe in clicker training. Read on…
I worked with a dog last night that had been exclusively trained using food and clickers by a local trainer. I’ve seen this kind of failure by clicker trainers over and over again. It is a 2 year old Labrador Retreiver. The owners had taken multiple classes. They were having trouble walking their dog on a leash; possessiveness over the food bowl and toys; and escalating confrontations between this dog and their other dog. The first issue to tackle was the leash walking. So, I first had them demonstrate what their dog was like on a leash. We went for a walk. The dog dragged this woman out of the home and down the street, barking. It was obvious it knew that a dog lived several houses down, because when we approached, it started lunging and barking as we approached, and sure enough, you could see that set off the neighbor’s dog barking. We went back to the front of their home and began a lesson. Within a half hour, you could walk that dog with one finger in the loop of the leash, all along the street, the dog was happy and relaxed, the barking had stopped, and for the first time, these folks could hang out and talk to someone in the cul de sac without their dog making an embarrassing display. I didn’t yell at the dog, beat it, cause blood to start squirting out of it’s neck, or inflict any pain. I use smart techniques and I understand dog behavior. It all changed in one lesson. The dog was happy and the owners were happy. Next week, we’ll deal with the possessiveness, and then any remaining conflicts between their dogs.
It’s many experiences, having to fix what clicker trainers have failed to do, which prompted me to write this article.
Clicker Training is one dog training method based upon a theory called Operant Conditioning. Those who believe that all behavior is explained by Operant Conditioning also believe:
1. All animals, including humans, are fixed-program machines. In other words, dog behavior is just a product of what we do to them, kind of like putting a dollar in a soda pop machine, pushing a button, and the can drops out of the bottom
2. There is no such thing as love, joy, sadness, fear or any other kind of emotion… BF Skinner said it was irrelevant for the study and practice of training or behavior modification; nor is the concept of “pack” relevant, and operant conditioning “advocates” are leading a movement to contend that leadership and pack order do not exist in the dog world
3. Humans are incapable of free will, internal motivation, reason or conscious thought — animals have internal motivations, but that is irrelevant — we are all machines, controlled entirely by our environment
4. Animals are just specimens to be poked, experimented on, caged, and eventually dissected for examination. There is no morality to Operant Conditioning
5. The Clicker Folks often contend, as a general rule, it is inhumane to ask a dog to inhibit a behavior, such as teaching a “leave it” command, or to teach a dog it MUST do something. Some go so far as to teach that it is harmful if you even touch a dog in training. Others combine clickers with traditional methods of correction… because clickers can’t do what even they were originally told they could do
6. That any other training approach is either cruel or completely ineffective
Do you believe this? Then there’s no point reading any further. If you don’t believe this, if you believe that animals have feelings and internal motivations and are more than just machines, then you have to reject the primary basis for the way many apply the learning theory of Operant Conditioning. You then also understand why in clinical settings, counselors are often prohibited from implementing operant conditioning programs on their patients… because there is no morality in Operant Conditioning. If an organism can be treated as a machine with no feelings, then there is no limit as to what you can do to that organism. Operant conditioning can be misused like any other system of behavior modification, and thus must be tempered by responsible, ethical… feeling… practitioners. (Funny thing, however… if you believe that humans have no feelings and are just machines… then how can they then exercise responsibility, ethics or feelings toward their subjects?) There are a number of court cases that should put any practitioner of what they consider operant conditioning on guard when implementing their methods, whether it is on humans or dogs.
AS you read this article, note that I am not discussing what I specifically do to deal with a variety of behavioral problems. I describe some of those things in my other articles at this website, and since I “sell” my expertise, I’m not going to give all what I know away for free. That would make no business sense. Thus, I’m using human examples. I could have used dog examples, but again, why give ideas to competitors, or those cheapskates who are looking for free advice instead of hiring a professional? Further, some types of dog situations are very dangerous, and I do not publish ideas on how to deal with dangerous dogs because of the liability risk if someone did what I said and they got hurt. Further, I DO understand and use the principles of Classical and Operant Conditioning in my work. That is NOT the same thing as saying that clicker training is the sole answer to training a dog. I would be a fool to say that I ignore the behavioral experiments, and training experiences, that have been accumulated over the years. However, watch a video on clicker training some time. Yes, you might see a nifty solution to a problem, and if that works, then go ahead and do that. On the other hand, you usually won’t see an analysis of why the dog is doing some undesirable behavior in the first place, and why maybe that nifty training program might just backfire.
Let me give you one dog example. I read in the news today about a 2 year old child that wandered into the backyard of her grandmother’s house, while grandmother was taking a shower. The dogs attacked and killed the 2 year old. Please tell me how clicker training would have prevented that tragedy. I can tell you what a good dog behaviorist would recommend to prevent such a tragedy, beyond the obvious answer of supervising toddlers around dogs.
Challenge One ********
Sam Basso: Remember all that stuff you learned about Pavlov’s dog and all that stuff you learned about B F Skinner’s pigeons in a box? Subsequent research proved much of that to be wrong.
Friend: How so ?
Sam Basso: They drew the wrong conclusions from the data. Once the experiments were done again they proved that the conclusions were wrong. It is too much to type here but I could explain it to you sometime. Classical conditioning and operant conditioning are being misused by many. They were never devised to train dogs, they were deduced from isolated experiments on a variety of animals. Neither Pavlov or Skinner ever trained a dog.
As the Riddler would say to Batman… “Puzzle me this”: Let’s say you wanted your wife to have sex with you on demand. So every time you wanted sex, you’d say a command, after she had sex with you as soon as you were done you clicked a clicker and gave her some chocolate. How well would that work out? Could you put the behavior on command? If you can understand that you can understand what I am saying
Looking at the overall concept, could I make a woman want to have sex with anybody just because they said a trained command word?
Friend: Well, no. Although… I could see people being like “I see the clicker and chocolate *waggle eyebrows*”
Sam Basso: Right but I’m talking about making a woman instantly have sex just like telling a dog to sit or lay down. That’s why behaviorism is not valid as a complete theory of everything. Behavior is more complex than the theory proposes. If the theories were correct, then they would work in all circumstances of whatever you were trying to get a dog to do or not do.
Oh how I wish they were wrong 😉
Friend: Hahahahahaha !!!!
Challenge Two ********
I’ll give you another good example of the limitations of operant conditioning…
Can you use operant conditioning to cause someone to laugh at the same joke again and again, minute after minute, hour after hour, day after day, year after year?
Nope. You can’t. There are other processes operating regarding behavior that are outside the realm of operant conditioning, but are still very important.
Challenge Three ********
I’ll use another human example… remember, if the theory universally works, then it should work on humans, too… Let’s say you have a child that is severely abused. You pick the abuse. Or a prisoner of war. Severely abused. This person develops a condition called dissociation… in other words, let’s say they develop a second personality. Britney develops a second personality she calls Amy. The adult prisoner of war, Jim develops a second personality he calls Fred.
Please tell me how you’d use clickers and treats as therapy for these two individuals. Let’s see, when Britney is talking to you, you click and treat her, and when Amy starts talking, you give her an electric shock with a dog collar… is that how it goes? When Jim is happy and doing well, you click and treat him, and when Fred starts talking, you put him in a time out… is that your plan for therapy?
********
Clicker training, as it is used by many, is based entirely a misreading of the theory of Operant Conditioning, even more narrowly than the theory upon which it is based. First of all, operant conditioning can only truly be done in a totally isolated enclosure, so that random responses can either be reinforced or punished. No one trains any dog like that. Most often, the concepts of punishment and negative reinforcement are also completely misrepresented and misunderstood. Instead, clicker training is often promoted as a utopian viewpoint that teaching and training should only be conducted using “positive reinforcement”, meaning using structured delivery of treats for randomly occurring behavior can build a trained dog… but it doesn’t work that way in reality. It also assumes that all acts generated can be controlled by whatever the humans put in the environment of the animal. Learning by reinforcement plays a major role in certain types of learning in all higher animals, including humans. However, it isn’t the only type of learning. Further, learning by purely positive reinforcement is even a smaller segment of the learning that can and does happen with all higher animals. Thus, we can see some amazing things taught through training by positive reinforcement, however, that isn’t always the answer when it comes to learning or behavior modification.
I can give example after example of behavioral and training situations where clicker training is ineffective. For example, go buy yourself a male Fila Brasiliero. I’ve owned two. This breed hates strangers, doesn’t trust them, doesn’t want them around, and will attack if provoked. They are 150 to 200 lb natural guard dogs. They typically prefer to be on home turf, aren’t super food motivated, aren’t much into fetching, are typically pretty dominant dogs (meaning don’t get one if you are a wimp or are the type to baby/ spoil a dog… do that, and breeds like this will be unmanageable… the last thing you need is a protection dog that will challenge you in your own home) and mostly just love those that they know. Yes, they are trainable and they are very affectionate and safe with the family, but it takes a lot of work. No amount of clicking and treating will get this dog to be accepting of intruders onto your property, however. Neither can you get the dog to accept a stranger staring them in the eye. The dog will tolerate it for a while, but if the staring continues, the dog is going to get agitated and start warning that it doesn’t like the challenge. Buy more than one (preferably a male and a female), then the effect will be even more intense. And it’s not just this breed. Lots of guarding breeds are this way, and you aren’t going to get a good result in dealing with their distrust and aggression using clickers and treats. They need leadership, proper training, supervision, containment, understanding of the breed, and so forth. Dogs like this have a vital role for those needing security (such as if a stalker is involved), but let’s not kid ourselves into thinking that working dogs of any breed are just pets. Further, no protection trained dogs are taught bite control using clickers. Answer me why that is.
Oh, I’ve heard about people claiming they can protection train a dog with clickers. Let’s just say that there is no need, and the uses would be very limited, indeed. You couldn’t go from beginning to end, training a dog to do the entire range of police work tasks and skills with clickers. Once again, I’m not going to discuss this in detail. First, because it isn’t my job to educate clicker trainers. Second, protection training is dangerous work, and I don’t want you doing it on your own.
Have you ever met a kid that was given everything… car, clothes, cell phone, food, etc. by their parents… and that kid was a complete jerk to everyone, irresponsible, and out of control? Don’t we all then wish that the parent would have exercised some discipline in their home to deal with their brat? Or have you ever been around a dog that was so “loved” that it was so obnoxious (or even dangerous) that you just wanted to not be around it? And hasn’t it been obvious that the dog needed some kind of leadership? (Oh… I should have mentioned, the concept of “leadership” or “pack leader” is rejected by those who teach Operant Conditioning / Clicker Training). What is obvious to the average person becomes lost by today’s “educated” elite in the field of behavior modification. This is a typical example where sometimes the most ignorant people are those who have the most education. They have learned so much that they now don’t know a thing.
Clicker trainers will claim it is inhumane to teach a dog to inhibit any behavior, only to teach them something “to do”. So, in their world, there is no place for a “leave it” command, or for teaching a dog that it must do something. I have a friend who was going to breed a dog. Short version of this… he took the female dog, in heat, with his intact male dog for a walk in the woods. Prior to the walk, he commanded his dog to leave the female alone, and his dog did, with one command (a well trained dog, by the way). When the arrived home, his dog was then given permission to mount the female and they did the deed. Can you do that with clicker training? Same dog was an attack dog, a dog that had actually been in combat with humans. The dog was once alerted to an intruder, and the dog started into attack mode and approximately 50 yards off, the handler called the dog back to him, the dog did. In another instance, this dog was provoked into an dog fight with a male Rottweiler. This dog had the Rottweiler down on the ground, still attacking, and the handler called his dog off the other dog and his dog came away from the fight. Can that be done with clicker training? I remember calling my own dog off an attack on a trespasser on my property, and within 10 feet of the man, my dog came back on one command to my side, and then I used a directed command to send my dog inside the house. And what about the dogs that I work with that are attacking one another in the home? I am currently working with 2 male siblings, about 17 months old, who are fighting with one another… a couple of 110 lb Rhodesian Ridgebacks. Is clicker training going to stop that? Or what about the 2 schnauzers I just finished with, where the older male would attack the younger male, and bite guests that came into the home? And what about teaching a dog that it must not potty in the home, which isn’t the same as teaching a dog to go outside, but instead teaching the dog that it must hold it’s urine and poop? I have also seen dominant working dogs, that will refuse to work. I remember a police type dog that would refuse to track. If it felt like it that day, it might do it, but at other times, would refuse. This was a test of wills with the handler. A lot of newfangled trainers discount that dominance exists at all anymore. The dog had very good appetite for food, healthy, etc. This business of using a “time out” didn’t work on this dog. And it doesn’t work on a lot of dogs like this, and some psychological pressure (I’m not talking abuse) is required to get them working.
Clicker Training is based upon a theory of learning called “operant conditioning” that is being misused. This might sound controversial, since clicker training is promoted as a “new” way to train dogs. Worse yet, operant conditioning is based upon the concept that all that matters is what an organism does in response to certain stimuli, primarily discrete consequences that happen outside in the animal’s environment. It ignores other valid research regarding the internal motivations, feelings and conscious experiences of the subject, in this case, the dog. Instead of reading just dog training books (written by people who want to sell you their stuff), you can verify my position on this by reading college textbooks and other research papers on behavior. We all know for a fact that dogs have internal motivations, feelings and a certain level of consciousness. Brain studies show that there is more going on than just input / output when it comes to learning, and a great deal of learning is a result of internal mechanisms that are not explained by Operant Conditioning. Yet with all we see, and with all this science, operant conditioning chooses to ignore those extremely important aspects of temperament, feelings or their effects on behavior. Thus, the concepts of aggression, pack order, and so forth are chopped away from the vocabulary of the operant conditioning theory, as if they don’t even exist. Further, operant conditioning ignores the sort of “unconscious” forces that motivate a dog, that go even beneath the immediate feelings the dogs are experiencing, or the training that the dog has received. Put in simple terms, in our own lives, we KNOW we have a personality with feelings and thoughts, and that sometimes those feelings and thoughts have been a result of the way we were brought up, the conflicts we have between choices, our frustrations, needs, and a whole lot of other things that sometimes require professional help (pastors, counselors, rabbis, psychiatrists, psychologists). Further, some things in our personalities are are not a result of just our upbringing. Two people brought up in the same household can turn out completely differently. We know there is a war at times between who we are and what others want us to do. Twin studies on humans indicate that approximately 40% of our behavior is genetically based and 60% is a result of our environment and experiences. Similarly, dogs are a lot more complicated than the kind of organic vending machine clicker trainers / operant conditioning promoters would have you believe. Dogs have internal motivations that have nothing to do with what we have trained them to do. A dog is going to behave like a dog, a Labrador Retriever is typically not going to behave like a Beagle. We selectively breed dogs to try and manipulate and control those internal motivations, but lifetime training and constant monitoring are necessary in order to control what a dog does, to prevent those internal motivations from causing the dog to do things that we don’t want them to do. Operant conditioning addresses, therefore, only a portion of what is going on with a dog, and clicker training based on positive reinforcement addresses even a smaller portion of what goes on with a dog… which is why it isn’t a balanced method of training and managing a dog’s behavior if it is the only method used.
Dogs are more complex than a light switch… flip the switch up, the light goes on… but it doesn’t work that way with animal behavior, especially higher animals such as mammals. The almost exclusive emphasis on overt behavior ignores what is going on inside the dog. There is a wide body of experimental and observational data and studies that discuss things that can’t be explained in the operant conditioning terminology. Most Skinnerian experiments and explanations of behavior are simplistic, and don’t address the wider range of things that most people deal with regarding their dogs beyond having a dog sit for food or jump through a hoop. Even the concept of a token economy (where human patients earn points, or stars or privileges for behaving in a desired manner) typically falls apart once the patient leaves the counseling facility. Skinner (the theorist who invented the Operant Conditioning model) believed that all behaviors were learned, which we know isn’t true, but it continues to color the entire world of operant conditioning theory. “Instinctive drift” (those internal motivations that define a species) has been observed in all animals, which can’t be explained by a purely Skinnerian approach, and even today’s theorists have to acknowledge that there is something wrong with the theory. Even Skinner eventually acknowledged, in the 1980’s, that Operant Conditioning was being replaced to a large degree by the proponents of cognitive behavioral theory. The majority of psychology practitioners had rejected his approach. Skinner was not very interested in overcoming internal emotional issues, such as anxiety, aggression, or other such internal states. Thus, I have had a number of conversations with operant conditioning fanatics who claim there is no such thing as aggression.
Dogs are more than just robots, responding to the external environment around them. Neither are people robots, otherwise there is no explanation for how people create music or art or new inventions, or self manage their time and direction in life. I have seen dogs do breed specific things, canine specific things, novel things, initiate things on their own. They are more than robots. And more is going on than the mechanical system proposed by Operant Conditioning. Further, there are social systems that are not addressed if clicker training is the primary tool of modifying behavior.
Therefore, this is why I don’t “believe” in Clicker Training, as many sell it. Their way is too mechanical as it is often used, too much about viewing a dog as a machine of inputs and outputs, has become an industry for a lot of people to make a lot of money based upon a sometimes harmful theory in the wrong hands, and why I have seen Clicker Trainers fail again and again when working with dogs. I DO have and use clickers from time to time, but that isn’t the same as what I’m referring to here.
I know many of you are fans of the theory of operant conditioning and clicker training, but as many of us found out when we took this into the real world and used it as dog trainers, we saw it didn’t always work as promised in many real world situations. We found out the limitations pretty quickly with real dogs. I use clicker training, I do use conditioned reinforcers, I have spoken with the main proponents of this kind of training, studied B. F. Skinner’s life and writings and theories, read the books, seen the videos, and I understand clicker training and operant conditioning very well. But, clicker training has many limitations, and so does the theory of operant conditioning. It is worthwhile to study operant conditioning, because many of the principles are valid and useful. But, it is foolish to rely completely upon operant conditioning as the only theory to explain what is going on and how to address any issues facing you, the owners, or the dog. That is NOT to say that I don’t use positive reinforcement in training. I am a big believer in using food, toys, petting and praise in training. I use primary reinforcers and conditioned reinforcers, and so on. I use them as much as possible and as much as is effective with every dog. However, there is more to training than using just these tools, and there is more to training any animal than the theory of operant conditioning or the practice of using clickers.
The most famous argument in favor of clicker training / operant conditioning is the one that says these trick trainers can train dolphins and killer whales to obey without using aversive corrections. What they don’t tell you is that THEY ARE ONLY WORKING WITH THE MOST SOCIABLE AND MALLEABLE MARINE SPECIES in these aquariums! They also use negative punishment, not just positive reinforcement. These techniques DO NOT work flawlessly with all marine life, and even these so-called expert trainers will admit that many sea mammals are not able to be managed safely in containment. They are too fearful or aggressive. These experts admit they can’t safely control the aggression or fearfulness with their techniques (while at the same time telling you that there is no such thing as aggression!). These noted experts specifically admit they are not willing to work with some species and certain individual animals. Further, it is not unusual for a wild animal to escape and never come back if the animal being trained isn’t caged in by a fish tank or otherwise tethered. This would be like a dog trainer claiming to the world they can train any dog to do anything so long as it is kept in a fenced yard, and the dog’s temperament is as a couch potato, food loving, mild mannered, field dog bred, neutered, male, adult Labrador Retriever. Dog trainers like me, and the owners of dogs that don’t fit this model (what kind of dog do YOU have?), don’t have the luxury of discarding certain breeds and individual dogs that don’t fit a utopian model of behavior. We have to work with what we get (and, by the way, not all Labs fit this model either).
Funny thing: even though these famous animal trainers admit that they can’t and won’t work with many species of marine mammals or certain individuals of their preferred species, this DOES NOT mean their disciples (all positive trainers) will tell you, while taking your money, that all positive reinforcement in training doesn’t always succeed with all types of dogs or other animals. Regarding dog training: What do they do with those dogs that are highly aggressive or fearful when they don’t fit within their program? The untold answer is a.) put them in kennels; b.) put them in harnesses, permanent barriers, or use electrical sensing devices; c.) Drug them; or d.) to recommend putting them to death. Huh? That is the best they can offer? Of course, they certainly can’t just turn them back into nature, like you can with a wild sea animal. They can’t just drop those dogs off in some forest or parking lot and figure it will all work out. There is a LOT of money being made by these so-called “all positive” trainers, promising results to unsuspecting dog owners. There is a lot of money. Go to any bookstore and look for these books and trainers; they promise the world, claiming to have invented a new “cruelty free” type of training (not telling you the greatest cruelty is what happens when they fail you and your dog, and your dog ends up in a veterinarian’s office with a needle in its forearm while you cry and the dog slowly drops into a quiet death). When people are making a lot of money, they won’t want to admit that they don’t know what they are doing, because then their money factory will shut down.
Any dog trainer, even a novice, can take a highly trainable, non-aggressive, food motivated, “no brainer” dog and get great training results. But, most dogs DO NOT fit that description. However, many dogs have internal motivations that will not be so easy to control using just a clicker training type of method.
The scary result of this is the creation of more and more highly insecure, spoiled, or dangerously aggressive dogs. Dogs need rules and structure in their lives; and yes, to learn to obey. When we coddle them, feeding their insecurities, not putting psychological pressure to be respectful, we cause them to grow up to be fearful, and in some cases, cause them to become fear biters or brats. When we coddle and spoil them, some dogs get very bratty. A spoiled dog is no fun to live with, is more destructive, is more likely to not be house trained, and is more likely to either be given away or abused once the owners are fed up. Lastly, some dogs, without proper leadership and corrections, will be made into dogs that are miserable to live with in the home and to take out in public. If you coddle some dogs, they will become dangerously aggressive. This is the untold danger of the “all positive” clicker training folks. And I’ve seen this trend more and more in my coaching. I’ve seen, over and over again, the INAPPROPRIATE over-use of positive reinforcement in training, causing or contributing to these problems.
This is illustrated with the trends in elephant training. Elephants are one of the most dangerous animals to work with. The clicker trainers have gotten a strong foothold in the zoo animal training business. Thus, old time elephant trainers, that used to take the necessary time and years to bond with, and gain leadership over, the elephants are being forced out. In their place, are operant conditioning trainers. What is the compromise being made? The elephants are no longer being worked directly by the trainers, and they are being put behind barriers where they can’t make physical contact with the trainers. The operant conditioning trainers have no good answers for the aggression, so they put up cement and metal barriers instead. They are also preaching that it is wrong to ever say “no” or ever correct any animal, claiming that it makes the animals more dangerous. You can’t have it both ways. You can’t claim your methods work when the only way you can use them is to confine animals behind concrete and steel. This is also happening in the dog world, where some trainers are saying that you shouldn’t even touch a dog in training. Yeah… they are real geniuses, huh?!
Operant conditioning is only a part of what is going on, a slice of the entire pie of what it means to be a dog. Like all theories, it explains some things very well, but it doesn’t explain all things very well. In addition, the operant conditioning folks even ignore their own theory, preferring to ignore the principles of negative reinforcement and punishment. They pretend that only half of the theory exists or is valid. Weird, huh? Talking with these folks is of like dealing with religious fanatics.
Here is an excerpt from an excellent article that explains one example (there are many others) why operant conditioning doesn’t always work in the real world. Clicker trainers, in my opinion, over sell their abilities:
A BRIEF HISTORICAL REVIEW OF ANIMAL BEHAVIOR STUDY
“It is evident that by the end of the 19th century, scientists who studied animal behavior in natural environments learned that the mechanical approach could not explain all behavior.
Behaviorism
During the middle of the 20th century, scientific thought again reverted to the mechanical approach and behaviorism reigned throughout America. The behaviorists ignored both genetic effects on behavior and the ability of animals to engage in flexible problem solving. The founder of behaviorism, J. B. Watson (1930), stated that differences in the environment can explain all differences in behavior. He did not believe that genetics had any effect on behavior. In The Behavior of Organisms] the psychologist B. F Skinner (1958) wrote that all behavior could be explained by the principles of stimulus-response and operant conditioning.
The first author visited with Dr. Skinner at Harvard University in 1968. Skinner responded to a question from her about the need for brain research by saying, “We don’t need to know about the brain because we have operant conditioning” (T. Grandin, personal communication, 1968). Operant conditioning uses food rewards and punishments to train animals and shape their behavior. In a simple Skinner box experiment, a rat can be trained to push a lever to obtain food when a green light turns on, or to push a lever very quickly to avoid a shock when a red light appears. The signal light is the “conditioned stimulus.” Rats and other animals can be trained to perform a complex sequence of behaviors by chaining together a series of simple operant responses. Skinner believed that even the most complex behaviors can be explained as a series of conditioned responses.
However, a rat’s behavior is very limited in a Skinner box. It’s a world with very little variation, and the rat has little opportunity to use its natural behaviors. It simply learns to push a lever to obtain food or prevent a shock. Skinnerian principles explain why a rat behaves a certain way in the sterile confines of a 30 x 30-cm Plexiglas box, but they don’t reveal much about the behavior of a rat in the local dump. Outside of the laboratory, a rat’s behavior is more complex.
Instincts versus Learning
Skinner’s influence on scientific thinking slowed a bit in 1961 following the publication of “The Misbehavior of Organisms” by Brelands and Brelands. This paper described how Skinnerian behavioral principles collided with instincts. The Brelands were trained Skinnerian behaviorists who attempted to apply the strict principles of operant conditioning to animals trained at fairs and carnivals. Ten years before this classic paper, the Brelands (1951) wrote, we are wholly affirmative and optimistic that principles derived from the laboratory can be applied to the extensive control of animal behavior under non laboratory condition]. However, by 1961, after training more than 6000 animals as diverse as reindeer, cockatoos, raccoons, porpoises, and whales for exhibition in zoos, natural history museums, department store displays, fair and trade convention exhibits, and television, the Brelands wrote a second article featured in the American Psychologist (1961), which stated, our backgrounds in behaviorism had not prepared us for the shock of some of our failures.
One of the failures occurred when the Brelands tried to teach chickens to stand quietly on a platform for 10 to 12 seconds before they received a food reward. The chickens would stand quietly on a platform in the beginning of training; however, once they learned to associate the platform with a food reward, half (50%) started scratching the platform, and another 25% developed other behaviors, such as pecking the platform. The Brelands salvaged this disaster by developing a wholly unplanned exhibit involving a chicken that turned on a juke box and danced. They first trained the chickens to pull a rubber loop which turned on some music. When the music started, the chickens would jump on the platform and start scratching and pecking until the food reward was delivered. This exhibit made use of the chicken’s instinctive food-getting behavior. The first author remembers as a teenager seeing a similar exhibit, at the Arizona State Fair, of a piano-playing chicken in a little red barn. The hen would peck the keys of a toy piano when a quarter was put in the slot and would stop when the food came down the chute. This exhibit also worked because it was similar to a Skinner box in the laboratory.
The Brelands experienced another classic failure when they tried to teach raccoons to put coins in a piggy bank. Because raccoons are adept at manipulating objects with their hands, this task was initially easy. As training progressed, however, the raccoons began to rub the coins before depositing them in the bank. This behavior was similar to the washing behavior raccoons do as instinctive food-getting behavior. The raccoons at first had difficulty letting go of the coin and would hold and rub it. However, when the Brelands introduced a second coin, the raccoons became almost impossible to train. Rubbing the coins together in a most miserly fashion] the raccoons got worse and worse as time went on. The Brelands concluded that the innate behaviors were suppressed during the early stages of training and sometimes long into the training, but as training progressed, instinctive food-getting behaviors gradually replaced the conditioned behavior. The animals were unable to override their instincts and thus a conflict between conditioned and instinctive behaviors occurred.” Genetics and the Behavior of Domestic Animals (Chapter One).,Academic Press 1998. San Diego, California. ISBN # 0-12295130-1 Behavioral Genetics and Animal Science,Temple Grandin and Mark J. Deesing,Colorado State University, Fort Collins, Colorado
The example with the chickens and raccoons involved a conflict between learning and a fixed action pattern (a type of instinctual behavior: An ethological term for a sequence of responses, usually but not necessarily produced by a releaser, the consistent patterning of which cannot be attributed to the operation of an operant chain. When the stimuli that elicit or set the occasion for a fixed action pattern are absent, their presentation, and thereby an opportunity to engage in the fixed action pattern, may serve as a reinforcer.).
While these are very clear examples of such conflicts, you will also find conflicts whenever you try to teach a dog something that it conflicts with the temperament of the dog. And you’ll have trouble training a dog when you remove the idea that you have to manipulate and manage the dog’s relationships with the family, other animals, and strangers. So, for example, say you want to train a dog to do a command-oriented behavior, but a drive or motivation prevents the dog from doing what your training is trying to accomplish. Thus, you can’t train every dog to be a police dog. The dog might have teeth and the body of a Doberman, but you can click and treat all day long, and use leashes and collars, but you can’t make a dog into something that its instincts are telling it to do exactly the opposite behavior. Similarly, if you have an aggressive police dog, you can’t train the dog to not be defensively aggressive when threatened. And much of protection training and protection dog handling can’t be done with clickers or with operant conditioning. You can’t take an animal that is highly reactive to sound, such as a hearing ear dog, and use operant conditioning to not be reactive to sound. You can also train a dolphin to do a bunch of tricks at Sea World, but there is no way that you can get a wild dolphin in the ocean to do those same tricks voluntarily because it can get its own fish and its desire to be with others of its species will interfere with your training. My beef with operant conditioning is that it treats dogs, or people, like machines. Input=> Output. Thus, the discrete behaviors you teach with clickers will often be negated in the real world when behavioral systems are activated to deal with those situations. Then, you have to use other methods to still get the desired discrete behavior in that situation.
I had a recent conversation with a college professor, a proponent of operant conditioning, who told me that they don’t use the term “aggression” or “pack”, nor do they believe in them as concepts. Strange, huh?. Really, weird. Especially considering you can see this stuff right before your very eyes if you watch dogs in groups. Have they ever read Lorenz? Just because you are a professor, and you decide to deny something exists to make it fit your theory, doesn’t mean you are dealing with reality. (Isn’t that what’s wrong with academics these days? Things have become so political and absolutist. It is like the Catholic Church persecuting Galileo for saying the earth revolved around the sun, except the new behavioral “church” is the Church of Extremists and The Politically Correct, ruled by the operant conditioning / “all positive” / clicker training “priests”, persecuting those who preach reality because it conflicts with this new church’s utopian ideals of how the world should work. These are also the same folks who won’t let a parent be a parent, and would rather let a kid run a household than teach them discipline, self respect, honor, and rules. [Oops, sorry. Back to dog training] There is a lot more to training a dog than operant conditioning, especially concerning the formation and motivations of the pack.
Operant conditioning theory used by most “clicker trainers” incorrectly assumes:
- All you have to deal with is observable behavior;
You can ignore the feelings and motivations of the dog and still get good results;
Genetics doesn’t play an important part in behavior;
Negative reinforcement, positive punishment, leadership, pack relationships and psychological pressure are not only irrelevant, they are abusive and dangerous;
You can ignore the different social systems that animals live within (packs, herds, schools, flocks, etc.), or think that positive reinforcement is the only way those social systems are maintained;
The only behavior that is important is voluntary behavior
Clicker Trainers take things even further by incorrectly assuming:
You can train all behaviors using positive reinforcement. WRONG!
Ever wonder what these so-called animal trainers do with the wild animals if a fight breaks out? Or if one develops a dangerous grudge against another or a human? You can bet that they use aversive techniques (we just aren’t told about it). Or, they get rid of the animals that don’t fit their programs. (Have you ever looked at the rules for the classes in which clickers are used? They only accept friendly dogs; however, not all dogs are friendly. So what happens to all those dogs that aren’t friendly? Are they to be put down because they don’t fit some utopian training method?) THE PURELY POSITIVE TRAINERS WILL ARGUE AGAINST CORRECTIONS, BUT WHEN THEIR METHODS FAIL, THEY WILL RECOMMEND THAT YOU PUT YOUR DOG TO DEATH! Does that make any sense?! What is worse, correcting a dog or putting the dog to death? Only a fanatic can’t figure that one out.
There are many dogs that will NOT obey if you don’t use corrections (read my article on Corrections… I am not referring to abuse, yelling, screaming, beating a dog, frying them with electric collars, etc… No one who loves dogs does that kind of thing) and psychological pressure in the training process. Some dogs are going to be dangerous if they aren’t given leadership. Some dogs will be killers without leadership, corrections and pressure. What kind of dog do you own? What kind of dog do you want? It is irresponsible to own a dog and not be able to control that dog.
I bet that the trainers at animal parks and zoos have to sign confidentiality agreements to prevent telling the public how often they are injured, or when they have had to use aversive techniques to deal with dangerous wild animals. In other words, DON’T BELIEVE EVERYTHING YOU HEAR ABOUT ALL POSITIVE / CLICKER TRAINING. It has its limitations. I have seen dogs that became such brats because of being allowed to choose whether to obey or not, that the dogs started biting people. I know of a specific case of a Doberman which was babied and just given positive reinforcements that recently bit someone on the street. I warned the owner that this was coming, and she ignored me. I recently worked with a 5 month old Welsh Terrier that was clicker trained for 3 months, and got so bratty, because it was never corrected by the trainer, that it seriously attacked the owner when he tried to take something away from the dog. They then hired me. I gave them a one hour lesson, and two weeks later all the biting had stopped, and all that was left was the dog occasionally putting its mouth on the owner. A couple of weeks more, and this problem stopped. Their “all positive” clicker trainer, a “professional”, didn’t know what to do about the biting. Now tell me, why is that? And why would you pay money to someone for them to wreck your dog like this? And what if this dog got older and had seriously attacked a neighbor’s dog, a child or a person? Would that have been better than just using proper leadership with the dog?
We all know that there is more going on inside a dog than what we see on the outside, and to ignore drives and emotions is a huge mistake when training a dog. I think Operant Conditioning is too mechanical, and ignores too much about the subject of the training. Yet, I do use the principles of operant conditioning, but operant conditioning theory is just a part of the entire spectrum of learning, and it doesn’t explain the entire learning process. I also use the principles of classical conditioning to deal with other training issues, and other training theories to help the PEOPLE involved (they are a big part of the equation; good dog training isn’t just about the dog) because they oftentimes better deal with other issues going on with the dog, such as aggression, anxiety and fearfulness and the people (such as their own personal issues and family dynamics).
What I do is not harsh nor inhumane. If a person walks along a path, and I tap them on the shoulder and say, “hey”, watch out! Have I done something harmful? If a person stubs their toe on a rock, are they permanently harmed psychologically? If a dog is focusing on something, and I command them to leave it alone, have I done something harmful? If my dog encounters an intruder, and gets in a full blown fight with them, yet my dog comes out the winner except for some bruises, is the dog psychologically harmed for life? Dogs and humans are more resilient to things than the all positive people would have you believe. Animals in general are not as fragile and neurotic as clicker trainers would portray, otherwise nothing would have survived this world. Are only the animals kept in cement pools, given fish for tricks according to their theory, the only psychologically healthy animals in the world, even more than those in the wild? Is that what their theory suggests? That nature is destroying animals in the wild because they have to learn not only good things, but sometimes things they shouldn’t do? Is the mother wolf harming her pup when she growls to tell them to stop suckling? Is the alpha male harming the other pack members when they get out of line? Are all these animals in the wild neurotic? What good trainers do to dogs doesn’t even approach the sternness of what happens to animals by their own kind. Something to consider.
Pack order is now the latest victim of the positive crowd. Agreed, dogs don’t usually form the more rigid types of packs that wolves do in captivity, but that doesn’t mean that pack related behaviors don’t exist in varying forms. Read more stuff, and you’ll see this going on. I’m very well aware of the trends these days in dog training. And I am very clear in my classes by what I define as being your dog’s leader. It has nothing to do with being abusive to a dog. And I disagree with the other notion that some all positive trainers assert, that dominance is all about access to resources. I’ve worked a lot of dominant dogs, and it is much more than this. If I am the one who puts the food bowl down on the floor, and controls when the dog gets to eat, does that make me the dog’s leader? No. Gaining a leadership role over a dog goes way beyond this kind of simplistic thinking. Dogs do not become aggressive when I train them (excluding the dogs that were already dangerous and aggressive before I even met them… then it is my job to help manage and contain and modify those feelings in the dogs). The dogs also don’t become fearful of me as I train them, either. I don’t purposely inflict pain, which is what the Operant Conditioning fanatics are implying when they say that there isn’t any humane training method but clicker training.
So, no, I don’t buy into the Clicker Training fad. And it is a fad. For some, it is a cult. They promise too much and get limited results, while taking your money. Funny thing, when these guys fail, I’m the one who gets the call and finally fixes the problems. The dog is happy. The owner is happy. No one was hurt. No cruelty. World peace hasn’t broken out. Life goes on.
So, Are There Useful Results From Clicker Training?
Yes, there are situations where clicker training is going to be useful.
As I’ve said before, I can’t save all dogs, but I can save most of them. And I don’t need clickers for companion obedience. Don’t feel guilty. You don’t need them either
FOLLOW UP: I’ve gotten criticism from the clicker folks regarding this article. The interesting thing is to see that a.) None provide any credentials proving their methods are better; b.) they don’t like the tone of the article – then again, a lot of people, when all they do is talk among those who only agree with them get offended when they run across someone that isn’t in lock step with them; c.) they try to attack me personally, as a “hater”, which is totally unfair and judgmental about someone whom they have never met; d.) can’t point to any specific dogs with significant accomplishments that have been clicker trained – hunting, police, obedience, tracking, etc… and just finding a specimen or two isn’t very convincing evidence… get enough dogs and sometimes you’ll find the genius dog that almost trains itself; e.) there is no point by point refutation of what I’ve said, supported by third party evidence. Clicker / Operant Conditioning advocates are like members of a cult. So, they get offended when you don’t agree with their terminology or method. And they will try to suppress any research or opinions that conflict with their orthodox beliefs. This even happens in the universities, not just in the dog training world. Just warning you if you go that route.
Instead of focusing on Operant Conditioning as a theory, we should be discussing a broader term which encompasses the Operant Conditioning theory, other learning processes, and then relating those processes to a specific species. And in circumstances where the theory doesn’t apply, we shouldn’t be forcing dogs to do things which the theory doesn’t address. Methods and techniques that work with one species won’t work with other species. And certain behaviors aren’t a result of, nor will be effectively changed by, associative learning. Further, we should be looking at the entire process of learning and behavior, not just on one aspect of behavior. When I’m working with a dog, there aren’t always associative learning solutions to the problem. For example, let’s say you have 8 dogs in the home, and they are fighting. Clickers and treats probably aren’t going to be effective to stop the fighting. Further, if you train enough dogs, you’ll see that what a dog “knows” and does is greater than their learning history. You don’t have to teach every little step of every behavior a dog does in order for a dog to do that behavior competently. You don’t have to train behaviors, like a clicker trainer would, by waiting for a dog to “offer” some random behavior, then reinforcing it, and each additional approximate step along the way. Thus, there are other processes going on that you can take advantage of… if you know what you are doing and have enough experience knowing how dogs figure things out and how they will respond to certain types of learning situations.
Answer this: do you believe that everything to know about behavior has been discovered, and is fully explained by the theory of operant conditioning? If you believe that, then you either think that a.) every behavior problem of every organism can be solved by operant conditioning; or b.) we’ve reached the limit of our learning and the degree to which behavior can be managed and controlled. I don’t believe either of these conclusions, and most likely, you don’t believe it, either. We all have that human relative, for example, which even the best psychologists can’t change regardless of how much we currently know about operant conditioning. We also run up against behavioral problems and training objectives for our dogs that aren’t completely addressed by operant conditioning, especially when we limit ourselves to using clickers. I believe that Operant conditioning theory has, for all practical purposes, reached its limitations. There are some folks who are so rigid, however, such little tin soldiers, that they can only see behavior from a reductionist viewpoint and entirely miss the intuitive Gestalt of animal learning which reveal the higher order systems in action. So, it makes them angry. Yet, anyone who is well educated in the topic of behavior knows that no new experiments are going to be made to greatly further our understanding of how behavior operates using this method. Behavior, not just learned behavior, cannot be completely explained by this theory. Which is why I say the theory is rigid and becoming obsolete, because people aren’t accepting what it can and can’t do. A grand synthesis of a number of other principles of behavior has to be proposed and tested. Emphasis needs to be placed on formulating this next level of understanding. My studies of behavior, and my practices of behavior modification, are focused upon that aim. I can tell you this: operant conditioning has become an abused theory… and that is a proven fact if you are willing to study real behavioral science. It was at its peak in the 1950’s, and lost its credibility after more research by scientists… but, you won’t see that if all you do is read dog training books, which are also over 50 years out of date.
There’s more to behavior than Operant Conditioning or Classical Conditioning. And there’s more to behavior than clicker training. I don’t “hate” operant conditioning as a theory. I’m saying that if you are basing your dog training solely on that theory, you are not going to be able to solve lots of behavioral problems, or training situations… and I consider you a novice who has no business teaching dog training classes to other people. You are not an expert. You are not ready to be a professional. You haven’t done your homework.
NOTE: I love this article, because it is a Troll magnet. It is interesting to see the kind of nutballs that write me on this topic. I don’t generally don’t post each Troll’s remarks because it ends up promoting their Narcissistic Personality Disorder, which would better be addressed in a mental hospital, with padded rooms, soft music, and jello for a snack. If we can help them put away their little plastic toys for a while, maybe we can give them back after they are cured of their obsession. Seriously… you’d be surprised at the angry messages I get on this article… They keep missing the point of this article: there is more to learning than operant conditioning; in many situations you’ll deal with when training most dogs, there’s not a need to use clickers, and it should be a choice of the trainer, not a cult obsession; that there are limitations to using conditioned reinforcers of all types; that their theory is causing behavioral problems when not being fully applied, and strict adherence to that method will fail so solve many behavioral problems. Of course, associative learning is real, but the theory is now being questioned and revised and new approaches are being developed because life is not a Skinner Box. It’s obvious with the examples in this article that there are many situations where a clicker trainer will be helpless if they stick to their method.
Here’s a recent email:
Dear Sam,
“I just wanted to write a quick word to you that I think your clicker training article is just about the best one I have ever read! As an animal control officer, I see the product of ”correctly done” clicker training every day. Not to mention I get sick of seeing people spending hundreds and hundreds of dollars on over-done dog management / confinement / gagdets to avoid discipline at all costs. I train horses as well. I have never met a clicker trained horse that wasn’t pushy, dangerous, and ”over excited” about getting food for performing stupid empty tricks. Anyway, I don’t want to suck up your time. Thank you for writing that intelligent article… dogs are mostly the product of their genetics, but the training and discipline that follows can really make or break the individual. I’d love to see someone clicker train a Border Collie not to display or feel herding behaviors- or clicker train an APBT to get along with every other dog it meets- or train an Ovcharka to be a therapy dog. It’s just sickening to be told I am being “cruel, abusive, and unnecessary” when clicker trainers have never actually seen me train a dog. Nor have they ever seen how well rounded my dog is (who isn’t brainwashed… and can actually make his own good and bad choices on a conscious level). It’s a fad. It will probably take a while to die, but it will. Eventually, there will be too many dangerous dogs due to lack of life structure. Dogs just aren’t STUPID ENOUGH for clicker training. Training is an art- it’s emotional. It’s not rocket science. It’s not a magical formula. Mother dogs don’t correct their puppies to be mean or give them PTSD…. they do it because that is how nature works. To imagine that an animal would actually hold a grudge against someone who gave it structure, correction, and encouragement is preposterous, unproven, and would result in the animal kingdom being too emotionally delicate to survive. Have a good one, Sam- and keep up the good work”
[PLEASE READ: Is Training A Dog With A Clicker Harmful? ]
Sam Basso is a professional dog trainer and behaviorist, in the Phoenix/ Scottsdale metropolitan area. He’s known for being fun, kind, intelligent, and humane. Sam Basso has a unique personal touch. He has appeared on his own TV show, been a guest radio expert, gives seminars, publishes a dog related blog, does rescue volunteering, and is active in promoting animal welfare and fair dog laws.