Positive reinforcement. Part 2. Understanding the pitfalls.

In part 1 of these series of articles on positive reinforcement, we looked at what it is exactly and why you would want to use it as you build the relationship with your dog.

In part 2 we’re going to look at some of the unintended pitfalls that can arise without you realising it.

So let’s get straight into it. What could possibly go wrong with using positive reinforcement? Your dog does something you like, you reinforce and therefore start to build an understanding that good things happen when that behaviour is repeated. Simple.

Or is it?

To understand why it might not be quite as straightforward as that, we need to start with ourselves and think carefully and clearly about what exactly is reinforcing to our dog.

To be successful at training any behaviour, whether that be shaping a brand new pup to empty out on cue in the correct place, as opposed to your living room, or negotiating a tunnel for the first time on the agility course, we start simple. First, we manipulate the situation to ensure the dog has the best chance of success. With our 'emptying out on cue' scenario, this means taking the time to understand when a puppy will most likely want to empty (after play, after food, after sleep for example) and - at these times - taking the puppy, on lead, to the place you would like them to empty. Staying with them until they do (do not cue them at this stage, they have no idea what you're talking about and you could well poison your eventual cue - more on this in an upcoming post) and strongly reinforcing when they successfully empty. There will always be unfortunate accidents which will require a quick clean up, but by consistently following those simple steps you will very quickly have a dog who chooses to empty in that spot because they are reinforced for doing so and what do we know? Dogs are more likely to repeat what is reinforcing.

With the tunnel scenario, we manipulate for success by starting with very short tunnels, positioning the dog as close to the tunnel entrance as possible so it can't miss and reinforcing strongly when the dog goes through. Once we have built their confidence with that, we can add some small challenges and build from there until they are ripping through tunnels from any angle.

But let's be clear, how we reinforce, when we reinforce and what we reinforce are equally as important as this initial shaping. And perhaps more importantly, we must understand that dogs are more than capable of teaching themselves behaviours if we inadvertently reinforce them.

So let's get into that.

Every dog is different in what they find most reinforcing. And it won't just be one thing. It's far too easy to think, 'Oh, my dog works for food,' and proceed to shape everything using food rewards. It may be true that your dog finds food highly reinforcing, but what about the myriad of other things it finds even more rewarding?

How about dashing around with other dogs in a field? How about swimming? It might be chasing a ball, chasing cars, runners or bikes. Or it could be sprinting through a tunnel or belting over an A frame.

You may notice that not all of those examples (chasing cars bikes and runners particularly) are desirable behaviours that you want to encourage. However, I'm sorry to be the bearer of bad news, but there is an extremely good chance that we have all at some point encouraged our dogs to continue with a behaviour we find undesirable because we haven't understood where that behaviour sits in our dog's reinforcement hierarchy and taken the necessary steps to reward an alternative way of behaving.

You may also wonder how on earth you might utilise these sorts of reinforcements in your training. Well, some, like chasing cars, you certainly won't want to, but others can certainly be used to build desirable behaviour if we do it correctly.

So where to start? At the beginning, of course. Here's a job for you. Take a sheet of paper and make a list of ten things your dog finds reinforcing. Be honest and truthful with yourself and your dog. Just because you might find it mortifying that your dog loves chasing joggers doesn't mean it isn't reinforcing and shouldn't go on the list. When you have your list, rank it from one to ten. Ten being the least reinforcing thing and one being the absolute most rewarding thing your dog can do. You know, the one that makes you say 'Oh, he'll do that all day if I let him.'

Now we have our list, what do we do with it? Well, first off, congratulations, we have just taken a massive step forward in understanding why our dog might be doing certain things. One of the biggies for example is pulling on a lead. Why do dogs constantly attempt to choke the air from their lungs and drag us around when they could just choose to walk nicely? The answer lies with what happens when you take the lead off. They get to play. That might be sprinting off through the long grass, it might be chasing after a ball you throw or it might be heading into the undergrowth for a good sniff at all the local - and not so local - peemail.

The bad news for us is that any and all of those things are far more reinforcing than walking nicely by your side. But in good news, lead pulling, like everything else can be fixed using positive reinforcement and by making our top ten list, we now have an armoury of reinforcements and rewards that we can use to shape it.

Yey!

Let's look at another very simple example from within agility.

We have dog who will not hold a start line. By that, I mean we position our dog before the first obstacle and ask them for a sit. We take a step forward and walk towards the first obstacle. We are about to give our first cue when a blur of eager fur and legs jets past us and launches over the first jump.

What we wanted, of course, was for the dog to wait for the cue we thought we had taught them before setting off. So why didn't they? And here's a clue, it's not because they are being deliberately naughty or stubborn or not paying attention, it's because the reward of doing agility has a far, far higher value than sitting and staring at a jump.

So how did we handle the situation? Well, by far the most common reaction for a dog breaking a start line is for the handler to think 'Ah well, that wasn't quite what I wanted, but they're off now so I may as well carry on.'

Maybe the dog doesn't set off sprinting, but they do the bum-scooch or they stick one paw forward and lower their head in anticipation. In this instance, the most common reaction is for the handler to call back to the dog with a 'No!' or 'Stay' or some other form of verbal correction. They then take a second step forward and notice another bum-scooch. They might correct again or, more usually, they will take that second edge forward as the cue for them to cue to the start because, at least, the dog didn't set off before you gave the instruction.

No harm done right?

Wrong..

By doing this, what we have done is reinforce them for breaking the start line and fundamentally changed the rules of the game. In both these examples, the dog got what it wanted and in both these cases it was the dog who made the decision when to start the agility run, not you. Therefore, don't be surprised the next time you line your dog up that the exact same thing happens? Don't forget the golden rule, dogs are more likely to repeat behaviours they find reinforcing.

Which brings us to how do we make sure we don't fall into these reinforcement traps? In a sentence, we have to understand why our dog is doing what they are doing and, most importantly, the part we are playing in reinforcing the dog to choose that particular behaviour over another so we can fix our behaviour

There, I've said it. All this is our fault. So let's take a deep breath, admit our flaws and start to fix it, which will be the subject of part three.

So let's summarise.

1. Make a hierarchy reinforcement list. And keep adding to it as you discover other ways your dog is reinforced.

2. Understand your part in the reinforcement process. Are you inadvertently reinforcing a behaviour you hadn't intended to reinforce or are you not reinforcing strongly enough a behaviour you want.

3. Don't blame the dog if we haven't taken the time to do one and two.

That's it for now. Next time we'll continue along our reinforcement journey and look at how to get it right.