In the world of dog training, there are two camps that seem to be at war -- pure click-and-treat dog trainers and everyone else.
This last faction is sometimes reduced to "Cesar Millan" by clicker trainers who are new to dog training, but it in fact includes Barbara Woodhouse, William Koehler, and 2,000 years of successful dog trainers on six continents, including the two people who are most responsible for the creation of clicker training, B.F. Skinner and Bob Bailey.
Now here's the funny thing -- Cesar Millan and click-and-treat trainers are doing the same thing!
What? How can that be?
Millan never seems to use a clicker, while the clicker trainers never seem to talk about being "calm and assertive."
But have you noticed what a clicker actually DOES to a new trainer?
It makes them "calm and assertive!"
Let's start with Cesar Millan.
When Millan talks about being "calm and assertive," what he is really talking about is making fewer signals and making clearer ones.
The average dog owner that goes to Millan is the typical "American nervous wreck" who is perpetually trying to do five things at once -- pet the dog, change channels on the TV, instruct the kids in the next room, and eat a slice of pizza, all while "teaching" one dog not to squabble with the other.
For the dog, it's a bit like listening to five different radios on five different stations all at once. What the hell is going on?!
When Millan tells a dog owner to be "calm and assertive" the first thing that happens is that the owner becomes self-aware.
What? I am not being calm and assertive? I thought I was!
Suddenly the dog owner stands a little straighter and moves around a little less, and when they do move now it is with a little more purpose.
The owner starts watching Millan, who may instruct him or her to simply walk ahead with confidence while ignoring the dog.
Suddenly, for the first time in weeks, the dog's owner is not sending mixed signals to the dog. In fact, he is not sending any signals at all!
Peace at last!
What does Millan do next?
As a general rule, he takes the dog for a walk, and on that walk clear consistent signals are sent with the use of a leash. The owner is asked to not look at the dog and not to talk to the dog. If the dog starts to acts up, a corrective "leash pop" is sent down the leash, perhaps with a small "Tsssst," a neutral sound made with the mouth, but there is no looking at the dog or talking to the dog.
Only ONE simple signal is sent, and that signal says "we could use a little less of that."
If you watch Millan, you will notice that he has exquisite timing and can read most dogs like a book, often signaling "we need less of that" at the start of a dog getting wound up or acting out.
And what happens? Most of the time, a quick miracle!
Now, let's look at clicker training.
What does a clicker do when put in the hands of a new wanna-be trainer?
Think about it.
When a trainer has a clicker in hand, and is focused on getting the noise timed exactly right, is the trainer flailing around with his or her hands?
Is he or she talking?
In fact, they are not supposed to be moving at all.
And in clicker training, it is the clicker that does the talking, not the human.
Is the clicker assertive? You bet!
The clicker sends just ONE clear signal -- a signal that says "we could use a little more of that."
So what's the difference between the proper use of a clicker and a well timed and properly delivered leash pop?
Not much, other than the obvious --- one signal is saying "give me a little more of that," and the other signal is saying "give me a little less of that.
The main difference is when the two signals are used, and here we come to the REAL difference between Cesar Millan and pure click-and-treat trainers.
At the beginning of every show, Millan tells the audience exactly what he does for a living:
"I re-habilitate dogs, I train people."
Listen to that carefully. Millan RE-habilitates dogs.
In short, his job is to correct what is broken. He does not habilitate; he RE-habilitates.
You do not call Cesar Millan to train a new puppy.
You do not call Cesar Millan to train a young dog for basic obedience or agility.
You call Cesar Millan when your dog is a mess and is acting out in a dangerous or disruptive way.
Millan's specialty is not writing on a blank slate -- it's correcting the confused scribble that has been laid down -- often for years -- on dogs that have never been exercised, are poorly socialized, are phobic, and are deeply confused about their role in the household.
Read that last line again. Millan's jobs is CORRECTING.
You cannot correct with positive training.
To correct, you need to use a combination of extinguishing behavior (i.e. completely ignoring some actions and reactions) while sending a clear message that you want to see a little less of whatever bad behavior is being presented.
You can do a lot with extinguishing, but it can be slow, and it does not work in every situation.
Does Millan give positive rewards? Of course! Watch him! He loves dogs and he sends positive signals all the time, but they are not always food-based and they are never heaped on for no reason, and never given to an overly excited dog.
Millan may shift his body to put psychological pressure on a dog, or shift his body to take pressure off. Removing pressure is a reward.
Millan may pat a dog on the chest or head, offer it a little play time or a favorite toy, or even slip it a small piece of food. Millan is flexible.
Millan is a great believer in simply walking the dog. Exercise with the dog is a type of reward (time with owner), a type of remedy (activity soothes anxiety), and also a recapitulation of the pack hierarchy (the owner is reinforced as the pack leader who initiates, leads and ends activity).
What about the click and treat trainer? Here too rewards are flexible, though things tend to be a bit more rigid on the front end when the dog is just starting out and learning that the clicker is a marker and proxy for rewards to come.
As training progresses, however, a click-and-treat trainer will move from click-and-treat to click-and-sometimes treat. Rewards may shift from purely food to a pat on the head, a rub on the chest, or a toss of a favorite toy.
As a command or trick is learned, rewards become less frequent, with several clicks-and-no-treats followed by a "jackpot" of several pieces of kibble delivered all at once. As training progresses further, and the dog is making fewer and fewer mistakes, the period between jackpots may get longer and longer as the dog learns to "work for wages".
In essence the dog is "playing the game," the same as someone might play a slot machine. If you got a reward every time you pulled the lever at a casino, it would be boring. It's the surprise of the "jackpot" that makes the inveterate gambler keep coming back for more. The same is true for the dog.
Cesar Millan is a balanced trainer -- he uses all three parts of operant conditioning. As a consequence, I have very little doubt he could train almost any animal put before him.
Can a pure positive trainer do the same?
In most cases yes, but not in every case.
Simple: If you put a pure positive trainer in front of an adult African lion in full charge, that clicker is not going to help much!
It turns out that a clicker is a great tool if an animal is young, or if it is a blank slate and does not have mayhem on its mind. But it is not much use if you need to fend off a charging lion.
It is also not much use if you need your dog to stop common self-reinforcing negative behaviors such as chasing deer or chronic (i.e. not ameliorated by excercise) barking in the back yard.
For those kinds of problems, few things work as well as an e-collar in the hands of someone who has actually read the directions.
I always find it amusing to hear pure click-and-treat trainers act as if learning how to use a slip collar is so complex only a rocket scientist could figure it out.
"So many people use a choke chain wrong," they sniff, while never teaching the correct way, which takes all of 30 seconds.
How about e-collars?
Here too we hear a lot of tut-tutting. "Yes, they can be a lot of use in the hands of a professional, but ...." the voice trails off.
The suggestion is clear -- you are an idiot.
No one like you can ever be trusted to actually read the instructions that came with an e-collar, or watch the CD-Rom included in the package.
Run a chainsaw? Sure. Raise a kid? Absolutely! Cook your own food? Of course. But you are too stupid and sadistic to actually figure out how to use an e-collar or a slip collar without killing your dog.
But of course you can learn how to train a dog with a clicker! Absolutely! And never mind that you are going to have to read a lot to learn how to do it, and maybe even watch a CD-Rom or two.
Now isn't that funny? If you decide to clicker train your dog, you are suddenly deemed to be a reader. Pick another training system, however, and you are suddenly a knuckle-dragging illiterate who will never be able to do it right.
Now to be fair, a clicker in the hands of someone who does not read a manual is not going to harm the dog much. It might get fat, it might get confused, and it might not get trained at all, but it will probably not be harmed.
But was a dog ever harmed by someone who read Barbara Woodhouse's book on dog training? Not that I can find!
And so, to recap...
- Rewards-based signaling is the core way of teaching young dogs or "blank slate" dogs that have no serious behavior issues, and it is the main way all dogs are trained to engage in new tricks.
- Aversion-based signaling is the core way of teaching adult dogs that have embraced bad habits to cease those habits.
- Both types of signaling require decent timing so that the signal (either a reward or an aversion) is delivered exactly when the activity occurs. The quality of the timing is, to some extent, the quality of the communication.
- Regardless of what method of training is used (balanced trainers use both methods), the success of the trainer will soar if he or she is calm, is not flailing around with voice, arms and legs, and is sending only ONE clear signal at a time.