Module 12 : How Dogs Learn

Similarities & Differences Between Humans & Dogs

In this module we will discuss how dogs learn and how dog trainers apply this knowledge when
changing dog behaviours. Specifically, we will cover the following points:

Similarities & Differences Between Humans & Dogs

The Humane Hierarchy Model Of Training & Learning

Natural Reinforcement

How Dogs Learn From Other Dogs

Personality Differences & Dominance In Dogs

How Dogs Learn – 5m10s

12.1 Similarities & Differences between Humans & Dogs

Compared to humans, dogs have a far superior sense of smell and their hearing is much more
sensitive.

Dogs are not as self-aware as humans and do not appear to possess the same capacity for abstract
thought and reasoning. A human may learn by consciously reflecting on their past experiences, but a
dog does not have this ability. However, there is good evidence that dogs hold mental
representations of how they should behave; you have probably observed guilty body language in a
dog that is behaving badly, even before it is admonished.

It is also important to note that dogs have a very limited attention span compared to humans. When
training a dog it is best to keep training sessions brief – five to ten minutes is plenty. Beyond this,
the dog is liable to become fatigued and distracted.

However, dogs and humans do share some similarities when it comes to learning. Both are capable
of learning from non-verbal cues and both can learn complex sequences of behaviour if the right
reinforcements are in place. Both will follow the Law of Effect, generally preferring to engage in
behaviours that offer some kind of benefit or pleasure, and both are susceptible to operant and
classical conditioning.

12.2 The Humane Hierarchy Model of Training & Learning

The Humane Hierarchy is a useful tool that highlights both how dogs learn and how they
should be trained.

Outlined by behaviour analyst Susan Friedman, the model describes six levels of behavioural
interventions, ranked from least to most invasive. Let's look at each level:

Level 1: Health, Nutrition and Physical Setting

Some undesirable behaviour is caused by injury or illness, so it is sensible to take a dog for a check-
up if it is exhibiting unusual behaviour. For instance, a dog may be unwilling to get in the car
because it has developed a hip injury, not because it is stubborn. At this level, a trainer will not use
any form of conditioning.

Level 2: Antecedent Arrangements

Assuming the dog is in good health the next step is to evaluate the dog's environment. A dog's
behaviour is enforced and triggered by its surroundings. Sometimes simple adjustments may solve
the problem entirely. For instance, if the dog barks incessantly when letters drop onto the doormat
then the owner could perhaps install an external letterbox instead. Again, at this stage in the
hierarchy, the dog is not subjected to conditioning.
Level 3: Positive Reinforcement

If it is not possible to make adjustments to the dog's surroundings then increasing the frequency of a
desirable behaviour by giving the dog a reward of some kind is the next best option. In other words,
the trainer will use positive reinforcement.

For example, if a dog is reluctant to come when called, making a special effort to reward the desired
behaviour if and when it does occur will increase the chance it will happen again in the future. If
they rarely settle down quietly on their bed, giving a treat when they are resting peacefully teaches
them that their behaviour will yield positive consequences.

Level 4: Differential Reinforcement of Alternative Behaviours

At this level, a trainer will reward desired alternative behaviours and remove any existing
reinforcers that maintain the problem behaviour.

For example, suppose a dog has developed the bad habit of jumping on its owner when the latter is
sitting on the sofa. A trainer using this approach would advise the owner to reward behaviour that is
incompatible with jumping up, such as sitting and standing quietly near the sofa. This is an example
of positive reinforcement.

This differs from the positive reinforcement used at Level 3 because it focuses not on simply
endorsing a good behaviour, but endorsing behaviour that is completely incompatible with unwanted
habits.

Level 5: Extinction, Negative Reinforcement and Negative Punishment

These strategies should only be used sparingly because there is a risk of distress to the dog, who
may become confused about what they should be doing. Extinguishing a behaviour entails the
complete removal of all reinforcement. For instance, if you leave a puppy to whine in a crate through
the night they may eventually learn to remain quiet, but this might cause them a lot of distress.

Negative punishment works by removing something the animal wants. For instance, if a dog jumps
up at its owner because it wants to play then the owner might leave the room for a few minutes. In
removing the possibility for attention or affection they are administering a negative punishment.
Note that a trainer would also draw on Level 4 techniques to encourage alternative behaviours.

During negative reinforcement, a trainer removes something from the situation – or from the dog –
in a manner that encourages the dog to repeat a behaviour in the future. For instance, if a dog wears
a choke collar it will experience pain if it pulls away. Only by obeying the owner's command to walk
to heel will the pain go away.

Level 6: Positive Punishment

Positive punishment entails the addition of an aversive stimulus to a situation. For instance,
shouting, hitting or administering an electric shock via a shock collar are all examples of positive
punishment.

Although interventions at all levels within the hierarchy will work at least some of the time, the fifth
and sixth levels are generally considered undesirable at best and deeply unethical at worst.

Punishment may be necessary from time to time; for example, if a dog is about to cause itself or
others serious harm or has not responded to any other training method – but less invasive tactics are
best.

Punishing a dog can undermine the trust between the dog and its owner. Research has also shown
that physical punishment prompted an aggressive response at least 25% of the time. Therefore, the
methods outlined at the top of the Humane Hierarchy may not even be effective in many cases.

When applying training methods remember that a dog has a short attention span and memory. You
must administer consequences, whether positive or negative, as quickly as possible. Otherwise, the
dog will become confused. In the worst-case scenario, you might inadvertently reward or punish the
wrong behaviour.

For example, suppose you come home from work and discover that your dog, who is waiting for you
by the door, has been chewing a cushion. Clearly, this is an undesirable behaviour. You may be
tempted to administer positive punishment in an attempt to correct the dog. However, unless you
know for a fact that the dog was chewing the cushion immediately before walking into the house, the
dog will not know why it is being punished. In fact, they may assume that they are being penalised
for waiting by the door.

12.3 Natural Reinforcement

Dogs can also reinforce their own behaviour in a way that makes them more likely to
repeat it in the future.

For instance, suppose a dog chooses to lie beside a fireplace on a cold winter's day. However, after a
while, they may start to feel hot, which causes a sense of discomfort. In response, the dog might
choose to move away from the fire. There is no human involvement or conditioning involved; this
kind of natural reinforcement is spontaneous and based on instinct. Because the dog removes a
negative stimulus from the environment this behaviour is properly termed natural negative
reinforcement.

Administering negative reinforcement – for instance, by pinching a dog's ear until it obeys a
command – should not be conflated with natural negative reinforcement. In the former, the trainer
controls access to a reinforcer and does not usually allow the dog to escape it; often they expose the
dog to the reinforcer many times in an attempt to make sure the new behaviour is entrenched.

Another example of natural reinforcement is that of a dog learning to steal food from a table. A dog
that realises the dining table 'offers' delicious food will try to jump up at future mealtimes to seize
food unless it is taught otherwise. This is an instance of positive natural reinforcement because the
dog has introduced something new (the food) to the scenario.

Understanding natural reinforcement is an important piece of the puzzle when trying to understand
why dogs develop problem behaviours. It makes little sense to say that dogs are intrinsically “bad”.
Often their behaviour results from a desire to fulfil their basic needs, such as a need for food or a
need for physical stimulation.

Five Simple Steps for Positive Reinforcement Training

Positive reinforcement training is known to be the best and most effective form of dog training.

Use these five strategies to get it right:

Use one word like sit, heel, come, quiet, stay etc. commands for each action. Also ensure that
all your family members use the same words.

Keep a supply of treats on you and give him a reward as soon as performs a desirable action.
This will help him connect the reward to the correct behaviour that earned him the reward.

Don't turn training sessions into a stressful experience. Use the opportunity to deepen the
bond with your dog.

Once your dog has learned desirable behaviours, wean off the food treats and use verbal
praise instead.

Use positive reinforcement as a form of ongoing training and establish trust and confidence
between you and your pet.

12.4 How Dogs Learn From Other Dogs

According to psychologist Stanley Coren, dogs can learn new behaviours by watching other
dogs.

As a general rule dogs like the company of other dogs. On an anecdotal level, those who work with
dogs report that it is easier to train a puppy if it is introduced to a household with a mature dog.

For example, a puppy will learn from a mature dog that it is normal to urinate in the garden, rather
than in the house. An owner will certainly have to take a proactive approach to house-training – they
cannot assume that a puppy will always follow the adult dog's lead – but it is likely that the former
will tag along with the latter outside and start to urinate in the same area.

Another example is teaching nervous puppies to descend staircases. Most young dogs are uncertain
how to handle steps. However, if they watch an older dog move confidently up and down a staircase
then they are likely to follow suit. Coren advises that the older dog serves as a model thus triggering
the puppy's innate desire to do what those around them are doing.

Of course this natural tendency may come with a downside; a dog can also serve to model
undesirable behaviour, such as jumping up on furniture. If you are addressing problematic behaviour
in a dog that lives with other dogs it is prudent to make sure you train both animals.

FACT

The average working life of a guide dog is eight to nine years.

guidedogs.org.uk

12.5 Personality Differences and Dominance in Dogs

Just like humans, each dog has its own personality and set of capabilities.

Some are relatively outgoing, whereas others prefer a quiet and sedate life. Different breeds exhibit
specific characteristics which become apparent during the training process. For example, English
Setters are working dogs known for their affectionate temperaments and intelligence. They are
generally eager to please their owners and are relatively easy to train. On the other hand, Afghan
Hounds tend to be more stubborn and independent.

There are mainly five different types of dog personalities.

Confident dogs: They know how to behave as a team player as well as lead the action where
they need to.

Shy or timid dogs: They dislike chaotic, loud and brash environments and may develop fear
or aggression if you force them to acclimatize in strange places. Calm praise, reward and
praise are effective techniques to use with shy dogs.

Independent Dogs: They usually bond very well with only one person and may require
specialized training to get them to obey commands. They may not socialize with other dogs
much and may be happy on their own.

Laidback, happy dogs: These dogs are enthusiastic, excited and happy most of the time. You
can need training in order to calm them down as they may like jumping on people. Typical
examples include Labrador and Golden retrievers.

Adaptable dogs: These are slightly different from happy dogs. They are eager to please and
much easier to train. They know how to be friendly without being exuberant.
What about dominance?

Contrary to popular belief, dogs as a species are not usually motivated by a desire to establish
dominance over their owners, trainers or other animals. Classic dominance theory states that every
dog wants to improve their social rank and they will resort to aggression if it will help them realise
their goal.

This view of dominance stems from old research into the behaviour of captive wolves. Researchers
noticed that wolves that lived in artificial packs – i.e. unrelated wolves that had to live together – did
sometimes fight to establish a dominance hierarchy. The most powerful wolves would ensure they
had the best access to food and the best mates.

Those interested in dog behaviour reasoned that because dogs are descended from wolves they must
have similar tendencies. However, more modern research of wolves living in the wild actually shows
that they prefer to live in cooperative family groups. In most cases, one pair mates and other wolves
help them rear their young. Wild wolves would rather leave a group than stay and fight others.

In addition, dogs have been domesticated for 60,000 years. During this time their behaviours have
become removed from those of wolves. When allowed to live freely, dogs will sometimes leave their
families. They may or may not be aggressive towards others but their behaviour depends on
numerous factors, such as prior relationships and experiences. Their social structures are flexible
and fluid.

What does this mean for trainers? In brief, research suggests that although some dogs may have
more assertive or even aggressive personalities we should not assume that they are driven to
compete with others for the status of “top dog.” You do not need to use harsh “dominant” training
methods to encourage desirable behaviour.

Assignment

Puppy Profiling

Time: 45+ minutes

Research into behaviour and suitability of guide dog puppies.

Download the worksheet below and complete.

Module Summary
 There are many differences between humans and dogs when it comes to learning, but a few
similarities.

The Humane Hierarchy model informs us how to train dogs, and also provides valuable
insights into how they learn.

The principles of operant conditioning underpin many popular methods used in dog training
methods.

Dog behaviour can be entrenched by natural reinforcement.

Dogs can learn behaviours from other dogs.

A dog's personality and aptitude for learning varies as a function of their breed and history.

The concept of dominance is misunderstood by many dog owners; modern dogs are not driven
by a desire to dominate others.

[Tweet "I just completed Module 12 of the Pet Psychology Diploma Course"]