How dogs learn

always-learning

Yes, your dog is always learning, whether you like it or not. But how do dogs actually learn in this human world of ours? The following is something shared on Make your Dog Smile’s Basic Manners Dog Training Class, and we hope it will help all of you to understand the world of dog learning! (Thank you also to dogtec!)

Dogs learn in two ways; by association (classical) i.e. emotional response and by consequence (operant) i.e. by doing.

Let’s look at learning by association from a human point of view:

We humans learn by association, too. For example, when we meet someone for the first time we come away with an association—either positive, negative, or neutral. If we really enjoyed our interaction with the person, we are likely to be really happy to see them again. If we found them to be difficult or argumentative and then we see them again, we might get that little pit of dread in our belly—we have formed a negative association with that person.

And how about the dogs then?

Dogs experience the world this way, too, only they rely on this learning far more than we do. They are constantly forming associations—safe, dangerous, neutral or good for me, bad for me, neutral. These associations inform the decisions dogs make and the reactions they have to various situations and stimuli.

Here is a dog comparison:

A common example of associative learning in dogs is their reaction to the site of a leash. Dogs love leashes! Pull out a leash and the average dog will jump of joy. This is because dogs have come to learn that leashes predict walks, and walks are fun so we love leashes. In other words, dogs associate leashes with fun walks.

The amazing thing is that we can manipulate dogs’ associations to things. For example, most dogs would find a chair to be of no consequence to them; they would form a neutral association to it. But we could teach a dog to LOVE this object.

We could place a screen between the dog and the object. Every time we lift the screen the dog gets treats until the screen goes back down. Pretty soon the dog figures out that the presence of the object is predicting the arrival of the treats and you have a dog that loves the chair.

We could also reverse this association; we could teach the dog to hate or fear the chair we mentioned earlier.So instead of giving treats every time we lift the screen from in front of the chair, we punish the dog by giving a leash jerk, shouting or in other way punishing the dog. What do you think is the association to the chair now? Yup, quite quickly the dog will learn to dislike / hate  and even fear the chair…

The implications of this are enormous. We have to recognise that what we do influences the associations the dog is making while in our presence. Here’s an important example: Say we are walking a dog and he has a reaction to seeing another dog that we don’t like. It could be just that he barks in excitement, for example, but we don’t like it. We shout, “No!” and give him a leash correction. This happens each time we see a dog. Pretty soon our dog’s reaction to other dogs is terrible—he barks and growls and lunges and snaps because we have built a negative association: dogs equal pain. We have taught the dog to dislike or fear other dogs. We actually have the potential to take a dog who either likes others or doesn’t yet know how he feels about them and make him fear aggressive, aggressive, or fearful around other dogs. This is the main drawback to using punishment —it has a lot of side effects due to learning by association or emotion. And not to forget that one of the negative associations is with the punisher, which can affect the bond between person and dog.

What might we do instead if a dog is having a reaction to another dog? Say that every time the dog we are walking sees another dog it starts growling and lunging? Our dog has a negative association with other dogs and we have to reverse it. Treats are a good way to do this. But what if our dog is so upset that he won’t take the treats? If we are  afraid of spiders and one is put right in front of our faces, or we are shut in a room with loads of spiders, it’s going to be hard for us to listen to any instructions to sit down and stop screaming. But if the spider is held 10 meters away, and only brought in for short periods of time, and maybe we are being distracted by some conversation or chocolate, things are going to go better for us; we’ll be able to hear when asked to take a seat and compose ourselves. We need to do the same things for dogs who are afraid of something
or upset by it—we need to desensitise using the Three D’s: distance, duration, and distraction. We move the dog farther away from the upsetting object, try to keep the situation brief, and distract with our cheerful voices and treats. Remember, we’re not rewarding the dog for his ugly display because he is too upset to control his behaviour. We are trying to affect his emotional state so that we can then ask for different behaviour.

And then there is the way that dogs learn by consequence. And to start of the human example:

We can tell a school-age child that we will take him out for an ice cream when we see him next week to celebrate his good report card. When he eats the ice cream, he will understand that he is being rewarded for grades he got a week ago, which he got because of work he did over the course of a period of months.

A dog could never understand this—it’s way beyond their ability to connect events over time like this. Dogs learn by consequence like we do, but for dogs the consequence has to be immediate; it must occur right on the heels of the action that caused it.

For example, say we lure a dog into a sit with our hand. Then we rummage around for the treat, trying to figure out where we put it. By the time we deliver the treat five seconds later, the impact is lost; the dog may not realise it got rewarded for sitting. In the five seconds between the sit and the treat, the dog sneezed, sniffed the ground, and looked left. All of a sudden there was a treat. As far as he’s concerned, he got it for looking left. We’ll eventually teach that dog to sit, but it’ll take a while. Or, we might end up with a dog who sits and looks left as a matter of course.

This is why we use a marker word “YES” (or clicker if you so wish) —this allows us to tell the dog the precise moment he won the treat. Once we’ve used the marker word we buy ourselves a few seconds to get the treat out of our pocket because the dog knows what it’s getting the treat for. The word YES is a reward marker—it marks the moment the reward was won. To teach the dog to know that the YES means a treat is coming we use learning by association—we pair the YES with treats. Every time the dog hears YES he gets a treat. Pretty soon the dog understands that YES means treat, that YES predicts a treat. So even when we’re working with learning by consequence associations are constantly being made.

There are two main concepts to take from these ideas:

• One, that dogs learn in two ways—by association/emotion and by consequence/doing.
• And two, that because of these two ways of learning, dogs see the world in two ways: What’s safe/good for me and what’s dangerous/bad, and what works and what doesn’t.

The safe/dangerous outlook on life comes from learning by association or emotional response. When a dog is punished for peeing on the carpet in front of us, they don’t learn inside/outside—they learn that it’s not safe to pee in front of us, but it is safe to pee when we’re not there.

The what works/ what doesn’t work outlook on life is from learning by consequence or by doing. A dog tries staring at the refrigerator. After a while he gives up and doesn’t bother trying again because staring at the fridge doesn’t seem to work; it never opens. Dogs also might try staring at their people at the dinner table. Every once in a while someone gives in and gives them a bite. Staring at people while they eat works, so dogs continue to do it.

The important point here is that dogs world is safe/dangerous and works/doesn’t work, NOT right/ wrong. Dogs do not have the capacity for those abstract thoughts. Dogs don’t do things we don’t like in order to get back at us or be stubborn or naughty. This is a myth. Dogs just do what’s safe and what works. That’s all. If a dog barks at us to throw the ball and we throw it, rest assured they’ll do that again. If we ignore the barking they’ll eventually give up and try something else. They’re not trying to be obnoxious, they’re just doing what works. If we ask a dog to sit and it doesn’t sit it’s not being stubborn, we just haven’t trained him well enough yet. So we have to be patient with them and be careful what we pay attention to and what we ignore, and the ways in which we do so.

Wishing all happy and joyful training!

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s