Myths and Fallacies About Dogs

In this article, I will discuss some of the problems encountered in the training process which arise from common myths, fallacies and handler errors.

“Dogs want to please.”

This is the biggest myth. If dogs wanted to please their owners, no-one would have any problems with their dogs, and this is obviously far from the truth!
Generally they do want to please – themselves. As we all do. The traditional approach to dog training is to make the dog do what you want. The more modern “motivational” methods, which I prefer, are based on motivating the dog to want to do what you want.

Unfortunately, “dogs want to please” becomes twisted into “dogs ought to please. I am displeased, therefore my dog is at fault.”

Or avoid getting into trouble?

Praise, petting, attention and company are important to dogs. Isolation is stressful and causes many behaviour problems. Communicating our emotional state also has a profound impact on our dogs, who are receptive to expressions of fear, anxiety, aggression or happiness on the part of their owners. However, it is a giant leap from that to the claim that “dogs work just to please” or that dogs are motivated solely by praise. Behaviour that is insufficiently rewarded starts to break down and go into extinction. Usually traditional trainers resort to “correction” (i.e. harsh voice, reprimands, jerks on the lead, physically shaking the dog) when praise turns out to be an insufficient motivation to guarantee the dog’s behaviour. The handler interprets this as deliberate disobedience. The dog may start to “obey” again, but the motivation has become avoidance of correction, rather than reward. The dog is not trying to please so much as trying to avoid getting into trouble.She knows she’s done wrong because …

She knows she’s done wrong because …

“She knows it’s wrong …” This is one of the most widely held beliefs about dog behaviour problems – and it is totally misconceived. I am always amazed when I ask people how they know that the dog “knows she’s done wrong” and they reply “well, when I shook the ruined shoe in her face and yelled at her ‘you bad dog, look what you’ve done!’ she cringed and looked guilty”. Of course she did. If you had brandished an shoe in her face and yelled at her “you bad dog, rhubarb, rhubarb” any other day of the week she would have done the same. What she knows is that she is getting into trouble, or more specifically that you are threatening her. She responds with a display of submissive body language and appeasement.

The problem is that your belief that your dog “knows” makes you feel enraged and can be used as a justification for punishment. You come home to a scene of devastation, and as you start to have a fit, Dogmatix slinks in looking guilty. In fact she is fearful of an angry outburst and is trying to pacify you. You then feel justified in punishing her, because “guilt” introduces a moral element, the mens rea, the criminal mind. Needless to say, it doesn’t help to stop destructive chewing.

What happens is this :

1. You go out, leaving Dogmatix in a state of anxiety.
2. She relieves this tension by having a good chew on something with your scent on it.
3. She feels a lot better. Nothing bad happens.
4. Some hours later, you come home, see the mess she has made and and have a fit.
5. Dogmatix responds with fearful or submissive behaviour.
6. Next time you go out, she is even more destructive, because as well as being anxious about you going out, she is beginning to be anxious about you coming home as well. She can’t win.
7. After a few experiences, she makes an association between your presence plus destroyed remains on the floor and her getting into trouble.
8. This is why she cringes – not because she “knows what she did was wrong” but because she is anticipating (correctly) that she will get into trouble.

Getting into trouble will not make her stop destroying things, because nothing bad or unpleasant happens when she is doing it. The punishment is associated with YOU, not with what she did several hours earlier.

Your dog should never associate your return with getting into trouble.  Any remains of destructive behaviour during the day should be quietly disposed of without comment – preferably without the dog watching.

Keep tempting articles out of the way. Be aware of what your dog might go for. Beware of spending the weekend gardening, with your dog for company. On Monday, when you are at work, the dog might spend the day imitating you, looking for whatever you found so interesting, seeking out your scent – in short, repotting your seedlings.

Knowing is doing – to hear is to obey

Traditional trainers and many inexperienced handlers stop rewarding their dogs because they have the attitude that the dog “knows” the exercise now, and should not need any reward. This assumes that “knowing” is “doing” – which it clearly is not. If I said “jump off a cliff”, would you do it? No. Why not – don’t you understand English? Obviously you choose to act because you are motivated – by fear, by self-preservation or by the prospect of achieving your sky-diver’s certificate.

Knowing and doing are the same thing. If I ask a class of new students to hold the lead with their right hand, maybe one third will do so, another third will put the lead into their left hand, and the rest will stand around dithering in confusion.

These people are intelligent adults who speak English and have voluntarily come to class to learn and are willing to follow instructions.

If it was a traditional dog training class, I would be bellowing “I said ‘right’!” I might even slap their left hand, grab the lead and shove it into their right hands, while glaring at them. I might turn to a fellow instructor and say “this student knows what right means. She is being deliberately defiant!”  My fellow instructor might say “She knows what it means. Make her do it. Don’t let her get away with it.”

Breed stereotypes

If I had a dollar for every time I’ve heard someone say “My dog is very … (insert characteristic)” I’d be rich. People often describe their dogs or their breed as :
[  ] very dominant
[  ] stubborn
[  ] hard to train, untrainable
[  ] stupid

People say things like :

“My dog is a … (insert breed). Malamutes (or any one of a dozen or more other popular breeds) are very dominant, you know.”
Sometimes they can be, but this should not be an excuse for giving up on the dog, or blaming the dog for failing to do what you have never taught it to do in the first place. Dominant dogs usually have a lot of initiative and working ability and will respond very well to training when you show appropriate leadership and motivate the dog to respond to you.

“My cocker/springer spaniel/setter etc. is so stupid! All she does is run around after birds!”
But she is a bird dog. That is what she was bred to do.
“My beagle is really stubborn. He never comes when I call him. All he wants to do is put his nose down and follow the smells on the ground.”

Again, this breed was developed to follow scent trails. Try to find a way to work with your dog’s abilities and instincts. For example teach your dog to find things for you.

“My Golden Retriever is driving us crazy. She picks everything up and carries things around in her mouth. She gets socks out of the laundry, she brings our shoes out of the bedroom, she takes all the children’s toys from their bedrooms, she runs around with a toilet roll in her mouth. How can we stop her?”

She is a retriever. She has a strong instinct to pick things up and carry them around, and especially to bring them back to “home base” which can mean you or her bed. To manage the problem when she is young, perhaps you should keep the bathroom, laundry and bedroom doors shut, and try not to leave shoes and socks lying around the house. Then give her a chance to use her instincts – teach her to bring in the newspaper, to teach her to fetch her own food bowl for dinner, to pack up her own toys, fetch her lead and bring you your slippers in the evening.
Funnily enough, people don’t say “my Kelpie or Border Collie is so stupid … all he wants to do is round up sheep.!” Somehow we recognise that these dogs show working ability and intelligence, and more people realise that these breeds need to do a lot of fetching balls and frisbees to stay sane as domestic pets. However, I did speak to someone who had a Border Collie who chased joggers and cyclsts in the park, which was a problem. “I thought these dogs were supposed to be smart” lamented the owner. Yes, but “smart” doesn’t extend to reading your mind and doing what you want.

Dogs learn by repetition

Many training instructors and books will tell you this. But it isn’t true. One of the most common handling errors is repeating commands. The dog learns to ignore commands while people repeatedly say “Sit, sit, sit, sit” with increasing volume, and eventually push the dog’s bottom down. I have even had a person then turn to me and “See, my dog knows what ‘sit’ means.”

I think the idea that dogs learn by repetition comes partly from the popular image of “Pavlov’s dogs”. There they were, immobilised in harnesses in the laboratory, repeatedly exposed to a bell and food. They learned “by association”. But this was classical conditioning, used to produce the reflex of salivation.
For training purposes, dogs are more likely to learn from the consequences of their actions. If they do something which leads to “a good result”, something pleasant or rewarding, they will tend to repeat that action. They can learn very quickly indeed, with hardly any repetition, if they are learning something highly emotionally charged and relevant to their needs. A dog will learn very quickly that pawing at the door handle makes the door open. A motivational trainer will therefore make each of the dog’s pleasures (walking, playing, eating, getting a pat, going out, coming in) contingent upon the dog producing that “open sesame” correct response.

This has more to do with operant conditioning, which was pioneered by B.F. Skinner. It seems that rats learn by trial and error, pressing levers or running around in mazes and getting a piece of cheese if they choose the right way.

I often wonder how different popular consciousness would have been if Pavlov had confined himself to producing eye blink reflexes in rabbits, and Skinner had used dogs to find out how natural behaviours can be elicited and shaped by the contingencies of reinforcement.

“My dog understands every word I say …”

Because we talk, we assume that dogs will learn what our words mean. However, we can be giving an entirely different message to the dog. Dogs that appear to understand every word we say are probably relying on familiar routines, contextual cues, body language and tone of voice. They will also tune in to a key word embedded in a lot of verbiage – as illustrated by the Larson cartoon “blah, blah, blah, blah, GINGER … blah, blah, blah, blah, GINGER …”.

The dog is being stubborn or defiant

The idea that a dog “knows” something, and when she or he doesn’t do as commanded it must be because the dog is defiant, stubborn, dominant or challenging the owner, is one of the hardest myths to let go of. Usually, it is because the dog doesn’t know it well enough yet, is confused, hasn’t learned to generalise the behaviour to new or more distracting situations, or because motivation has broken down. Learning is difficult. With the best will in the world, we never get things 100% correct, especially in the early stages. Yet we expect this of our dogs.

So let’s look at some of the complexities of the training process, where handler error and miscommunication is rife.

Training glitches

Lack of clarity of communication in training takes many forms.

Inappropriate use of inducive movements

People who are inexperienced in using lure hand movements typically make these mistakes :

One is that when they have their pup’s interest and hold food just in front of the pup’s nose, they guide the pup into the wrong position. For example, when trying to lure a down from a sit, they move their hand too far away from the pup, inducing a stand instead.
Another is that as the pup follows the food lure (which could be above the pup’s head and to the right), they start to chase the pup’s nose, rather than guiding the pup into position, until the handler is doing ABC logos in the air, and the pup’s head is weaving all over the place, going anywhere but down. Pity the ABC has stopped using that promo. It would have made a cute one.

The result is that the pup becomes discouraged by lack of success in getting the food, and stops following the lure hand movement. The handler tries again, the pup sits and stares at the food without moving, then finally decides to give it another go. As the pup lowers his or her head, the handler instantly whips the food away, towards the ground, in the attempt to induce a down. The effect is that the pup is further trained to stop following the lure. Removing a piece of food when the pup moves towards it is, in operant conditioning terms, negative punishment. The best way to teach a pup not to approach food is to remove the food if the pup moves. It is an excellent method, but it is not a way to lure a down.

Using words (so-called “commands”) when the dog does not understand their meaning

There is a huge difference between training in the sense of “the teaching process”, and training in the sense of the finished product – “how a trained dog behaves”. Until modern methods based on “learning theory” came along, there was far too little emphasis on how to induce behaviour in the early stages of teaching (when the dog obviously doesn’t know it), or how to produce that behaviour reliably (when the dog is in the process of learning it).

This may seem obvious to instructors, but it is not at all obvious to inexperienced dog owners. If I say to my dog “lie down” and he does it, this is not training so much as the result of previous training.  All I have to do now is make that sure what he has learned is maintained by occasional reinforcement and not unravelled by contradictory messages.

This creates the illusion in the observer that if they say “lie down” their dog will do it. When it doesn’t work, the obvious thing is to repeat it, turn up the volume a bit.

You have to teach it before you can use it.

Overuse of words and extraneous vocalisations (aka padding)

My favorite command is “rhubarb”. It’s very useful because it can mean almost anything. The favorite command of people coming to classes (before I get to reprogram them) is “C’mon”. This also means anything (and everything). Clearly humans need to vocalise while training e.g. “C’mon, c’mon, c’mon, c’mon, c’mon … over!” when trying to entice a dog over a jump. For the dog, movement and body language are more important.

Conflicting words and body language

People often use body language which conflicts with their words and intentions. For example :
•    If a dog approaches and starts to jump up, many people will lean back and raise their arms – their body language is signalling the dog to “come on up” – while their voice is saying “get down”.
Instead, depending on how well you know the dog and how friendly it seems, I suggest standing side on, keeping your arms close to your body, then maybe putting a hand out at the dog’s nose height (not on top of his or her head) for the dog to sniff and make contact without having to jump up.
•    Most people if they pat a dog will make rapid hand movements around the dog’s head and face. This stimulates the dog, especially puppies to chase and mouthe person’s the hand.  “How do I stop him from biting?” a person will ask, while making vigorous, agitating movements around their dog’s head.
A good start would be not to cause the reaction in the first place. Stop moving your hand. Clench your fist and wait. Most dogs will lose interest. Them calmly and slowly stroke your dog on the neck, shoulder or chest.

Conflicting words and hand signals

It is typical of the learning process that you cannot walk and chew gum at the same time. This applies to handlers as well as to their pups. Veteran behaviour analyst Bob Bailey has repeatedly emphasised that training is a physical skill, involving co-ordination and timing.

I like to remind people, when they find it is harder than it looks, how they felt when they first learned to drive. Left foot on this pedal, left hand on the gears (that’s up – down is for reverse), right foot on the other pedal (no the other other pedal, that one’s the throttle), right hand on the indicator (no that’s the windscreen wiper), you need to put your hands back on the steering wheel to turn the corner, bang your head on the steering wheel in despair – lucky that’s where the horn is. You’ll need to warn the drivers who assumed that when you put the left indicator on it meant you were going to turn left …

However, everyone made it to class today, probably on automatic pilot.
So, we know that dogs go through various stages in the learning process:
•    just doing the action, because they have been lured, guided or reinforced, but at first they are not really aware of what they are doing
•    then you start to see some deliberate or intentional movement
•    then you can introduce a cue (such as the word “drop”)

At this point I like to remind people that their puppy is not dropping because they said “drop”, just because you induced the behaviour just after saying the word. This is like attaching a verbal label to the action – this word means this action.

•    after a while the pup will learn the meaning of the word, or at least know it to some extent
•    if you are using lure-reward methods, you should also at this stage get rid of the lure, although you still need to reward after the action and you probably retain an abbreviated lure movement as a hand signal

This is where both handler and dog suffer from the strain of being asked to walk and chew gum at the same time. The handler has to say a word, and use a hand signal. Being human, the handler focusses on the word, and all awareness of body language flies out the window. The handler is just as likely to say “drop” and give a “sit” signal.

The dog is having similar difficulties. Having been asked to shift his or her focus from the food to the action and now to the signal, the dog is being asked to attend to the word rather than to the body language or hand signal. A dog’s first language is body language. But not even a native speaker knows that an upward hand signal mean “drop” just because you said the word. The dog follows the the hand signal and sits, or if the body language is really indistinct, stands there looking puzzled.

We have a cross-species communication problem.

Even when dogs know a voice command (or cue) very well, and respond reliably to voice alone, they will usually follow a hand signal rather than the voice if the two are in conflict. Of course a dog can be trained to respond to voice alone and ignore the handler’s body language, which is essential for some working dogs, but I am talking about the average pet dog who will probably pick up a bit of both.

To illustrate the point I occasionally demonstrated with my old German Shepherd, River. She would sit, drop or stand with voice alone, but when I deliberately said “sit” and gave a drop signal, she dropped.  I only ever did it two or three times. I thought she could tolerate the odd handler error without her training unravelling completely. But when such errors abound in the training process, it is amazing how the dogs manage to learn.

Using the same word for different actions

For example, many people use “drop” for the down position and “drop” meaning let go of the ball. This is potentially confusing for the dog, but handlers can usually get away with it, largely because the dog is responding to contextual cues and body language, rather than the voice as such. That’s why we don’t often see a dog respond by lying down with the ball in its mouth.

Using the same vocabulary for formal and informal levels of control

I believe that incorporating “levels of control” into training is essential. My informal level of training is what you need  for practical everyday life control. Formal training involves a higher level of precision and attentiveness. It is used less often in everyday life, but is developed further for competition obedience.

Teaching “stay” as a formal obedience command, and then using it in an informal context, meaning “mill around at the front door but don’t follow me to the letter box” will erode your formal “stay”. It is better to use the informal “you wait there” at the front door, or when you open the car door.

Meaning it sometimes but not others

People tend to overuse commands. It’s bad enough repeating words ineffectually when you mean it, let alone when you don’t really mean it. I often see people halfheartedly calling their dog, and shrugging when the dog ignores them. I also see people asking their dog to sit or drop, then doing nothing if the dog simply wanders off. If it’s OK for your dog to wander off, why bother asking him or her to sit? If you say it, mean it, and think about how to handle any unwanted response.

I am not saying that you have to force your dog to sit. It is perfectly valid in a training session to reward good responses and give no reward for an unwanted response. You still mean what you say: sit means “sit and you get a reward, don’t sit and you miss out”. This is not the same as indicating that sit means “do your own thing, I don’t care one way or the other”.
Cues used by the dog – often not understood by the handler
A dog may be relying on all sorts of cues that the handler is unaware of, such as:
•    Handler’s body posture
•    Location
•    Familiar routines
•    Contextual cues
There is nothing wrong with relying on these cues, as long as you are aware of them. You can make them work for you. The dog can use them to distinguish between the show ring, the obedience ring and everyday life, where different behaviours and different levels of control are needed. However, it can work against you if the dog discriminates on the basis of cues that the handler is not taking into account. For example, the dog comes when called everywhere in the park – except approaching the exit, where you normally put the lead on and go home. In this location, coming when called is “punished” rather than rewarded.
It is surprising how much dogs rely on subtle cues, such as a slight tug on the lead, the merest touch of a finger on the rear, raising your eyebrows or clearing your throat after giving a command. If these are systematically eliminated, you can be left with a dog who no longer understands what “sit” means, a point amply illustrated Ian Dunbar’s well-known routine.

These should all be considered before you decide that the dog is being stubborn or defiant.

The great “knowledge” trap

•    My dog has done it a few times, therefore he or she “knows”.
One of the most damaging ideas in dog training is the assumption that the dog “knows” when really the dog is at some point along the learning curve. A great deal of harm can be done by punishing the dog for incomplete understanding or for baulking at obstacles encountered in the learning process. Even more damage can be done if the trainer expresses frustration and anger.

Many stages in the training process lead to glitches in the dog’s understanding that can be mistaken for deliberate disobedience.
I place a lot of emphasis on the importance of teaching, as all too often dogs are blamed for glitches in the training process. However, learning the meaning of a command, signal or cue is only the beginning. The greater part of training lies in establishing the long-term motivation for the dog to do it.