About Positive Dog Training

The world of dog training is changing. "Traditional" training by compulsion, based on the assumption that a dog needs to be dominated to be made to do what we want, is being replaced by a more modern approach. This approach is driven by an enhanced understanding of how dogs (and many other species) learn, and what makes them increase or decrease the frequency of a behaviour. Positive reinforcement (i.e. anything that will increase a desired behaviour by making it rewarding for the animal to perfom it) is now being used by more and more dog trainers as a dog friendly and more effective alternative to the traditional approach to dog training.

At the same time as our understanding of animal learning improved, the assumption that dogs need to be dominated to be happy and obedient was also debunked as a myth. 

View the RSPCA's position on dominance based training.
View the APDT's (Association of Professional Dog Trainers) position on dominance based training.
View the AVSAB's (American Veterinary Society of Animal Behaviour) position on dominance training. 

The expectation that a dog should work for us just because we own him, or that he will only work for us if he knows we are "the alpha", means that we spend a lot of time being upset with the dog when it doesn't do what we want. 
Once we let go of that expectation, we can get on with what we actually need to do:  Just train the dog.

Dogs are actually quite easy to train. They have no hidden agenda: they simply do what works for them. 
They work out what they need to do to:
- make good things happen, or stop them from going away
- make bad things go away, or stop them from happening.

In "traditional" training, the leash jerk on the choke collar is a "bad thing" that happens after the dog does something we don't want it to do. The dog will work out over time what it needs to do to avoid this bad thing from happening. In the process, it will also work out that when it hears our "nice voice" (the verbal praise we are allowed to use in this type of training), no bad thing is about to occur. Over time, this will teach him what is safe to do.
This training method exploits the law of learning that a behaviour that has negative consequences for the dog, will reduce and eventually go away.

By contrast, "positive" training aims to reinforce the dog (ie make a good thing happen) when he does the right thing. We help the dog understand what we want him to do by either initially showing him what to do (e.g. by luring with food in our hand, not by pushing him into position), by rewarding a behaviour he does on his own (like sitting down), or by rewarding a small step towards the behaviour we have in mind, gradually cranking up the requirements until we get to where we want to be ("shaping"). 
This training method also uses the law of learning, in this case that a behaviour increases in frequency if it has positive consequences for the dog.

So, both methods, if applied with good timing, consistency and sufficient repetition, will work.
The traditional method produces a dog that works to avoid punishment.
The positive method produces a dog that works to gain rewards.

The problem with traditional training

Dogs have been trained to the highest obedience competition levels with this method, so there is no doubt it can work. However, to work, it requires two things: a dog that can psychologically and physically cope with the discomfort and potential pain inflicted during training, and an owner who uses this technique expertly in a way that causes quick learning, as opposed to a long drawn out (and so more painful) learning period.

The biggest advantage of the traditional method is that it is relatively easier to understand and execute than the positive method: for most of us, it seems easier to identify what we don't want and react to that with a form of reprimand or punishment, than to focus on what is desirable and reward any step in the right direction.

However, there are potentially significant side effects to traditional training methods which use aversives:

• Dogs unable to cope with the discomfort or pain associated with training can become either uncooperative ("shut down"), fearful, defensive (eg barking at owner) or even aggressive. This is then interpreted as trying to dominate the owner and further punished, often leading to an escalation of the problem behaviour.
• The bond between the owner and dog can be negatively impacted, as the dog associates its owner with unpleasant experiences.

• Frustration and anger on the behalf of the owner can lead to injury of the dog, if these emotions are vented when the dog is on a choke chain. It does happen: I've been there myself, so I know what it is like.
  
• The dog might get used to ("habituate" to) the level of discomfort associated with leash jerks, requiring more pain to be inflicted to be effective in training.
• Dog owners often don't understand how to use the choke collar the right way, ie by delivering short "leash pops", never letting the dog put constant tension on the leash (and thereby reducing airflow and potentially causing damage to his windpipe). This is evident in the large number of dogs constantly pulling at the end of a leash despite their choke collars, effectively strangling themselves in the process.

"You can find pet dogs and obedience show ring competitors from both training styles that are happy, reliable, willing workers. You can find dogs from both training styles that are poorly trained and out of control. But you’re likely to see more dogs in a compulsion-based class who grudgingly comply with commands or look bored or disgruntled than in a positive reinforcement class, where enthusiasm usually abounds among two-and-four-legged students alike. More importantly, methods that utilize coercion, force and intimidation have a significantly higher likelihood of creating behaviors such as learned helplessness, in which the dog simply shuts down, and aggression, in which the dog fights back." Pat Miller, Do-Over-Dogs

For more information including videos, have a look at Dr. Sophia Yin's extensive discussion of punishment/dominance based versus positive reinforcement training.

The high number of cross over trainers speaks for itself.

A large number of positive reinforcement trainers "crossed over" from traditional methods, never to look back again.  Pat Miller, Pam Dennison, Mark Spector and our very own David Weston, founder of The Kintala Club, are some of the well known of these trainers. More

All of these trainers were training dogs successfully the traditional way, earning the highest obedience titles. Each of them had an experience with a particular dog, which made them question their methods, and subsequently discover the alternative way of training dogs by using positive reinforcement. They went on to be just as successful, or more so, using this method, and felt resoundingly it was a superior and more humane way to train. The trainers above, and many others, went on to publish popular books and slowly managed to spread the word about this different method of training. Positive Training is now on its way to becoming mainstream in the US. Cesar Millan's seemingly contradictory popularity may be the backlash any major change brings with it. 

From puppy to adversary

When we take a dog into our lives, most of us, I suspect, are looking for that special bond that can exist between dogs and humans. Of all the baby domestic animals, puppies are the most people oriented. They seem to come with an innate desire to be with us humans. They happily bounce up to anyone, wriggling their whole body in excitement. Everyone's expectations are high when the puppy is finally brought home. 
Then, at puppy school, we are told that we must show the puppy who is boss so it won't become a nightmare to be with when it is older. That it will only be happy if it knows its rank, or otherwise it will try to boss us around. 
We are told we must never let it on our bed, not let it follow us around the house, and lock it up in the garage at night so it learns to be alone, learns that it has no power, that we are the ones who control its life.
When it mouths us, or picks up something it's not meant to, we must tell it off with a firm voice, or even a growl. If it doesn't stop, it must be shaken by the scruff of its neck, because "that's what a mother dog would do".
And so the dream ends. We push aside our instinct to be gentle with the puppy, and treat it as we are instructed. We become used to being harsh with it. The puppy learns that bad things can come from us. It learns what it has to do to avoid our anger. 
We are now more adversaries than friends: to show the puppy its place, it must always be told off when it does the wrong thing. Be rolled on its back to learn to be submissive. We find ourselves getting angry at the puppy more and more often, because it does so many wrong things, even though we believe it should know better.
As it grows older, the puppy no longer comes to us happily. It starts to be more interested in the world around it, starts pulling on the leash to get to this world faster.
We go back to dog school for help. We put a choke chain around its neck. This is how we communicate with our dog, the dog we wanted to bond with: we give it a jerk to its neck with a metal chain if it does what we don't want it to do. The best it now gets is our happy voice if it happens to have guessed right what we want. Once again, the dog works out what it needs to do to to avoid the punishment of the leash jerks. So, the training is successful. 
It's not how you thought it would be. But that's what you have to do to a dog, don't you?

Don't end up in this trap: learn how to train your puppy the truly positive way, and enjoy life with your best friend! 

Delta Dogz - Smart Training!

• Private Training
• Behaviour Consulting

---

Find out more about the training services we offer.

Common misconceptions about positive reinforcement training

The dog will only work for the food, not for me.
Many people seem to be put off positive training because it involves giving the dog small bits of food as a reinforcement for doing the right thing, because somehow this means that the dog does not do it "for them". If the dog works for them, then equally no choke chain or any other aversive should be necessary. However training can generally only be achieved by using either reinforcement or aversives. 
For some reason, the use of aversives in dog training meets a more general public acceptance than the use of food, despite the demonstrated effectiveness (and reduced wear and tear on the dog) of using food in the context of positive reinforcement training. Positive training however is more likely to result in the dog actually working for you in the long run: over time training in itself, and by extension you, will become a rewarding experience the dog will want to seek out.

I will have to carry food with me all the time to get my dog to do what I want.
Food is just one reward we use in positive reinforcement training. Anything the dog wants and likes can be used, such as throwing a ball, a game of tug, access to other dogs, or patting. Small bits of food are very useful in training specific behaviours such as sit or drop, as well as tricks. Good trainers will phase out the use of food as a lure (as opposed to a reward) as soon as possible, so the dog does not become reliant on seeing the food to do the required behaviour. Food as a reward can be reduced and replaced with other rewards as the dog becomes more confident in his behaviour, or you can start asking for more "bang for the buck", ie you ask the dog to do more before you give him his reward. 
When I first started with positive training, trying to avoid having to have food with me as soon as possible was high on my list too. But as I gained more experience, and packing some treats became a routine, I stopped being so hung up about it. As your relationship to your dog changes, you will probably find you won't mind as much "paying" him for his work, at least sometimes. 

My dog is not interested in food, so I can't train him this way.
Some dogs are definitely more food oriented than others. However there is almost certainly some food your dog will like, so it's a matter of finding out what tickles his fancy. Roast chicken pieces for example are highly valued by most dogs. 
Food is also not the only positive reinforcement we can use, although it is a very convenient one for initially training a behaviour. Ball crazy dogs leave even roast chicken alone if only they can chase a ball: we can exploit this by teaching him what he has to do first (eg a sit) before he gets to do what he wants.
So it's a matter of finding out what your dog finds rewarding, and building that into the training approach.

Positive training means I can't do anything but ignore unwanted behaviours, because I am not allowed to punish the dog. 
This argument if often trotted out by those who only have a shallow understanding of positive reinforcement training (thinking it equates to feeding the dog biscuits when it does the right thing, and ignoring everything else), and who want to make a case for the use of aversives. 
Skilled positive reinforcement trainers have a wide range of techniques to deal with unwanted behaviours, depending on the context. 
In general, when using a positive training approach we show the dog what we want it to do by rewarding the right behaviour. When training a new behaviour such as a sit or fetch etc., any mistake (ie wrong guess of what to do) by the dog is simply ignored. If the training plan is set up correctly, the dog will work out by itself, via trial and error, what it is that will earn him the reward, and will start to offer that behaviour more often. In this context, unrewarded behaviour will slowly disappear.
However, self-rewarding behaviours such as jumping up, digging or chasing will not cease just because we ignore them. The inherent rewards in these behaviours also usually beats the value of any food reward we are reasonably able to offer the dog. 
The main ways of dealing with self-rewarding unwanted behaviours are to

• prevent them from happening ("managing" the situation), eg  by not allowing unsupervised access to the part of the backyard you don't want the dog to dig in;
• teach the dog what we want it to do instead. In the case of jumping up, we would teach the dog to sit or at least keep their feet on the ground in order to get a pat. Here's a great YouTube Video by kikopup illustrating this beautifully.
• teach the dog that its access to what it wants to do is contingent on doing what we want it to do first (also called the "Premack principle"), eg we let the dog greet another dog after he has walked loosely on the leash or heeled for a trained amount of time. 

Dogs do need to be taught how to behave in an acceptable manner, not because they otherwise want to lead the pack, but because life with them is too hard if they don't. Punishment however is never required to achieve that end.

For a positive way of interrupting unwanted behaviours, see this excellent video by Emily Larlham (kikopup, Dogmantics):

If I train this way, my dog will just work for whoever has food, he won't have any loyalty to me.
At off leash parks, I often get mobbed by dogs trying to get at the food I have with me. The dogs I get hassled by are dogs that never get any treats. Dogs that are trained with food as a reward have internalised that they get treats by doing something, as opposed to trying to get them by brute force. If someone does have nice smelling treats, my dogs might try and sit in front of them to see if that earns them a treat. Or they might try performing some of the tricks they know. But they don't try brute force.
Initially I tried to teach the dogs I got hassled by to at least sit to get a treat. It worked very quickly, they sat, and sat again ("throwing sits" in positive training jargon), due to the never experienced joy of getting something for what they were doing, completely ignoring their owners as a consequence. I had to stop this approach, because these dogs' owners would get quite annoyed with me and their dogs for their perceived disloyalty.  "Oh, you've got treats" they scoff.
Make yourself the source of all good things to your dog, and he will never dream of going off with anyone else.

My dog will get fat from all the training treats.
Yes, he will get fat if you feed him treats AND all his meals as well. So make training rewards part of his daily food intake. Food rewards should be very small pieces of food anyway. The majority of your food rewards should be lean and healthy (but obviously tasty) treats, such as cooked chicken. Standard commercial "treats" often contain sugar, artificial flavours, colours and preservatives, and should only be used in moderation. If you feed your dog kibble as his main meals, use some or all of that kibble as training treats instead if your dog accepts it. Usually kibble is not enticing enough in more distracting environments, or if a new behaviour is being trained, but you can use it in your home (spiced up with something more interesting) if your dog accepts it happily, or once a behaviour is well understood.

Some breeds must be taught who is boss.
Some breeds, such as Alaskan Malamutes, have a reputation of being "independent" and supposedly will "lead the pack" unless handled with a firm hand, ensuring they consider themselves the bottom of the pack. Interestingly, websites that espouse this view also state how intelligent this particular breed is. There are plenty of examples of positively trained dogs of this breed - ANY breed dog can be trained the positive way. You may have to be on your toes more with an intelligent, energetic dog, and be ready to shape the behaviour you want during all your interactions. Also, importantly, a working breed like a Malamute requires much more mental and physical stimulation than many other breeds. Supposed "dominance" behaviour may simply be caused by being chronically understimulated. Dogs in this state of mind are also near impossible to train, no matter by which method. So yes, the breed will determine certain traits, and you should certainly think carefully before getting any type of working breed. But NO dog of any breed needs to be battered into submission. Just train the dog with positive reinforcement - and look after its physical and mental well being. 

Too scared to play

Recently at our local off leash park, I saw a beautiful German Shorthaired Pointer. I had met its owner before, and knew she was a dedicated (traditional) obedience trainer. The dog was clearly very well trained, and under full control by the owner. But whenever another dog approached, it would become fearful and try its hardest to ignore it, paying even more attention to the owner. [This was different behaviour than a dog fearful of other dogs would display.] Even with the owner's encouragement, it would not engage in play. 
In the more advanced levels of traditional training, dogs are punished (by a leash pop, or harsher methods depending on the school) if they stop focusing on the owner. This dog had clearly internalised this, and was simply too scared to pay attention to anything but its owner. It had lost its ability (and desire) to interact with its own species. I don't know about you, but to me, this is not what I want. Whilst we all would like to be able to gain control over our dogs when we need to, most of us would also like to be able to let our dogs interact with other dogs, and have some fun with their own kind. 
This dog's behaviour shows one of the possible side effects traditional training can have.