Saturday, 24 March 2007

Aggression in Dogs

Today I was asked a question regarding the best way to manage a dog that displays dominance-related aggression towards people. My first thought when asked this question was how they could be certain that the aggression stemmed from a dominance issue. Aggressive behaviour can have a whole range of causes, and the reason for aggression really does need to be established before action is taken.

Ideally, all cases of aggression should be seen in person by a qualified, experienced behaviourist - each case needs to be treated individually, especially where aggressive behaviour is concerned, and so this is not to be considered a comprehensive guide to managing aggression. However, a few simple rules apply:
  • If a dog is properly raised, and managed carefully around puberty, there should be no issues around who is the dominant one in the relationship. However, in some circumstances aggression problems can still arise and need to be tackled as soon as possible.
  • The dog is behaving completely naturally for a dog - he doesn't understand that what he is doing is undesirable. He is not a "nasty" or "bad" dog, just a misguided one.
  • Try to avoid giving the dog any opportunity to get it wrong. For example, if he tries to bite people as they enter the house, simply arrange for him to be in another area when they come in. Look at each cirumstance, and work out a way for him to get it right every time.
  • Reward, reward, reward! Even the slightest success must be pointed out to him as the right thing to do. If last week he attacked your Auntie Nellie, but this week he just looked at her, throw him a party! The more positive experiences he has, the more positive he will feel towards the situation.
  • And on that note, DO NOT physically punish him or attempt to intimidate him. Not only does this create a negative experience which damages the dog-owner bond, it could also trigger an aggressive outburst. The only form of punishment that can be used in this case is negative punishment - the removing of something he finds desirable - but even this should be carefully considered, as aggression may arise if the dog perceives you as taking something that is rightfully his.

As you can see, the basic principles of training still apply. Positive reinforcement and negative punishment are a powerfully effective team when used properly, but when issues of safety are concerned, should only be attempted by experienced people. If in doubt, call in your friendly neighbourhood behaviourist!

Friday, 16 March 2007

Clicker Training



Clicker training has become one of the most-used animal training methods, and I'm a big fan!

Pictured here is the Kong Multi-Clicker, which is the type I use, but there are lots of different clickers out there, and I wouldn't particularly advocate one over the others. The purpose of a clicker is simply to make a clear, distinctive noise immediately it is activated.

And why would you want something to do that? Well, to explain that we have to get into "old-school" classical conditioning. Remember Pavlov and his dogs? Pavlov discovered that by ringing a bell and then feeding the dogs, the dogs made an association between the bell and the food - they salivated (and had other physiological responses) when they heard the bell.

This is what we want from our animals - to build an association between the clicker and a reward. This is possible with any animal that can hear, and likes to eat - not just dogs! A simple session of repeated click-treat is normally sufficient to build that association, and to get the animal looking for its reward whenever it hears the click. Once that association is established, a world of training possibilities opens up.

Operant conditioning applies from then on. When the animal does something you like, give a click and reward the animal. When it does it again, click and reward again, and so on. When the behaviour is being performed regularly and reliably, you can build in a cue (a command) for that behaviour, by giving the cue just as the animal is about to do it.

To make that clearer, let's use an Example Dog (and let's call him ED for short), and say that you want to train him to sit when he's told. Here's a step-by-step guide:
  • ED loves hot dog sausages, so make sure you have plenty of sliced-up sausage to use as rewards
  • Let ED know that you have a bit of sausage in your hand, so his attention is fixed on it
  • Raise your hand over ED's head so that he's looking up
  • ED finds this position awkward, so he'll naturally lower his bottom to the ground in order to get a good view of the reward-bearing hand
  • As soon as ED sits, click and treat
  • Don't worry if he tries jumping up and mugging you! Just don't give him the sausage, and try again
  • When ED sits as you raise your hand over his head, start saying "sit" as you do it
  • Repeat this a few times and he'll now build associations between the word "sit", him sitting, the click and the treat
  • Before you know it, ED has learned to sit when he's told

See? The purpose of the clicker is to mark the EXACT behaviour you're rewarding him for. It helps to clear up any confusion over whether you were rewarding him for wagging his tail, for whining, for licking his lips.. he knows for certain that you made that click noise because he sat. And most dogs will soon start to wonder what else they can do to make you click!

Training Equipment

So you want to train your animal. What equipment will you need? Obviously it depends on the species, and what you're training it to do.



Here, you can see me training an ex-racehorse, using Parelli Natural Horsemanship to introduce her to the idea of jumping, as she had been a flat-racer.

The halter, the line, the stick, the cavalleto (little fence) - they are all specifically designed for use with horses, and I wouldn't dream of skimping on them. If you're training a horse, there's no way round it - you need specialist equipment. Using makeshift tools can be dangerous, for you and the horse.

But what about smaller animals? Is there a cheaper, hassle-free alternative to buying lots of expensive training equipment for, let's say, a dog? Absolutely! The only limit to what you can use is your imagination, and of course safety - as long as it's not going to hurt your dog in any way, you can probably use it in your training.

Keeping on the theme of jumping, imagine you've seen flyball on Crufts, and are longing to give it a go with your dog - but you don't have any hurdles, and those flyball boxes look very expensive. Improvise! Try making hurdles out of whatever you have lying around the garden, and to stand in for a flyball box, you could have a mat and a friend. Teach your dog to jump the hurdles and touch the mat, which your friend will take as a cue to throw the ball in the air - dog catches it and returns to you over the hurdles. There you go - your dog now knows the idea of flyball, and when you finally manage to find a flyball group in your area, they will be amazed by how quickly (s)he gets the hang of it!

Use your imagination, be creative but always safe.. and have fun with your training!

Tuesday, 13 March 2007

Punishment

So what's my problem with punishing an animal when it does something wrong?

I actually DON'T have a problem with it, provided the punishment takes an appropriate form. To explain that, I'll need to tell you about the different types of reinforcement and punishment.

  • Positive reinforcement is adding something to the situation that the animal finds good. For example, giving a cat a food reward for coming when called.
  • Negative reinforcement is taking away something that the animal finds bad. For example, ceasing to use the cattle prod when the bull moves forward.
  • Positive punishment is adding something that the animal finds bad. For example, hitting a horse when it refuses a fence.
  • Negative punishment is taking away something that the animal finds good. For example, putting the ball away when the dog starts barking at you to throw it.

In modern animal training, positive reinforcement and negative punishment are the preferred tools of the trainer. This allows the animal to always focus on good things, and how best to earn them.

Using things that the animal finds bad can cause stress, fear and aggression towards the trainer. There is also the issue of escalation - if you smack your dog for jumping up, but he doesn't stop, what do you do to him next? This is a scary situation, best avoided altogether.

In my opinion, the only possible justification for using positive punishment is in an emergency life-or-death situation. As an example, a dog that chases livestock may be cured of this behaviour through the use of a citronella collar, or even a shock collar. But what if it associates the punishment with the livestock, and becomes MORE aggressive to them as a result?

There is no substitute for good socialisation and reward-based training.

Basic Principles of Animal Training

Some people make training dogs (or any animal) seem like such a difficult thing. It's really not! I work on the principles of Operant Conditioning:
  • If the animal performs a behaviour, and something good happens as a result, it's likely to do it again.
  • If nothing good happens as a result, it's less likely to do it again.
  • If something bad happens as a result, it's quite unlikely to do it again.. BUT punishment has its own set of complex side-effects. I'll write more about that in another post.

Pretty straightforward, huh? The only tricky bit is working out exactly what your particular animal perceives as a good thing. It could be something as obvious as a food reward, or a favourite toy.. but sometimes it's a little more complicated.

For example, take a barking dog. You see it everywhere - the dog is barking, the owner's shouting at it, the dog continues barking and all chaos ensues. Rewind. WHY is the dog barking? It could be fear, excitement, stress, aggression, or even just wanting attention. In which situation will shouting at the dog solve the problem? None of them. So the trick is reading the dog to decide why it's performing that behaviour, and in the context of that motivation, decide what the dog would perceive as a good thing.

Sometimes, as in the case of fear and stress, it's necessary to remove the source before the dog can learn because it has entered an entirely emotion-driven state. This is an unfortunate position to get into, as the dog can perceive the removal of the scary/stressful thing as a reward for its behaviour, so perhaps in future it would be wise to avoid such situations until you can introduce them in a more controlled manner. The same could be said for aggression and excitement, depending on the level being exhibited.

Always be aware of what triggers your animal's behaviour, and ensure that you have given the animal the tools to cope with these situations before it gets "thrown in at the deep end". It makes life easier for everyone!

Assuming the situation you and your animal find yourselves in is a controlled, unemotional one, the rules are simple. If it's doing the right thing, reward it. If it's not doing the right thing, don't reward it. This creates a totally positive experience for the animal, encouraging it to try harder to earn its rewards. And who doesn't want an animal that actively wants to do the right thing?

Monday, 12 March 2007

Introduction

Well, this is my first post on my new blog, so I guess I should tell you a bit about who's writing this.

Hi, I'm Mismorphic, and I have a degree in Applied Animal Behaviour and Training. I'd like this to be a resource for people seeking advice about their animals, and I'm willing to help however I can. My key areas of expertise are:

  • Animal training
  • Domestic animal behaviour
  • Domestic animal care (NOT veterinary issues - that's what vets are for!)

It may take me a while to post all the things I'd like, so if you have a question you think I may be able to answer, but you don't see anything here, please email me and I'll reply as soon as possible.