Monthly Archives: August 2007
This is a test article to ensure that submission and editing of articles via the front end works. Usually you write a brief introductory teaser to the main article body here…….
…and then write or paste the main body of the article here. (If you paste it, make sure it is unformatted)
Use the buttons at the top to save, apply or cancel. Use the insert image and insert page break buttons below the editing box to add images to the article.
Use the images tab at the bottom to select, order and format images (they must already have been uploaded to the server).
This part of the site contains administrative data about the site and information about me, and will probably never be finished…..
For various technical reasons, the easiest way for me to write new articles is online, using the content management system’s editor (and thanks to high speed mobile broadband, it’s cheap and easy to do this!).
Once I get past a certain stage, I publish the article in progress so that I can see how it is formatting and how readable it is by using the site front end. I indicate articles that I am still working on by starting them with the phrase (In Progress), including the brackets.
This means that if you find an article which starts like this, please don’t write to me to tell me that it doesn’t make sense, looks peculiar, has empty pages or peters out in the middle of nowhere. I already know.
However, as I’m more than capable of starting seventeen articles and finishing two, I’ve made a rule for myself that says I’m not allowed to have more than one In Progress item on the site at any one time, so if you find two or more, feel free to send me a note and rap me over the knuckles.
I’ve also made myself a rule that says I can’t leave an In Progress item untouched for more than three days. I either have to continue working on it or take it off the site for the time being. So if you see something In Progress that looks as though it’s gathering dust, feel free to rap me over the knuckles for that, too!
All this is positive punishment, of course. Alternatively, you could send me a rave email to tell me how fabulous the article is when I finish it and take the (In Progress) marker off. The rate of reinforcement might be a bit low, though
Rescue or shelter dogs which have been abused by humans can be remarkably difficult to clicker train – apparently! In fact, unless they are exceptionally traumatised, most of them are only too grateful for food and attention, and are capable of becoming wonderful workers and even of competing, if they have good working drive. But there are several things which can go wrong up front, and which may make the dog appear to be reluctant to work. Although there is no single solution for anxiety, and no single approach to rehabilitation, here is a case study of how clicker training can apparently fail, and how to remedy the situation.
First and foremost, dogs who have been abused have good reason to be afraid of humans. Such dogs may be reluctant to approach humans, accept affection, or work, and may even have a tendency to snap or snarl at people. This needs to be seen as the result of real and justifiable fear, and definitely not as disobedience or naughtiness. The protocol for resolving this is desensitisation, not discipline.
Secondly, abused dogs, when finally placed in a rescue home where they are treated with love and kindness, often bond exceptionally closely with their new family. But this usually does not generalise to other people, and in particular, it may not generalise to someone giving a training class or to other people in the class.
In other words, your rescue dog may become extremely attached to you in a very short space of time, but he may nevertheless be very frightened of the trainer, the behaviourist, people in class who try to pet him, or the vet, for example. There is no information in his mind to help him distinguish between people who will hurt him and people who won’t.
And thirdly, if he has been hit or kicked, he may be able to tolerate people moving around him going about their normal business, but find the approach of a hand or foot very threatening, and become anxious.
So the first rule of conduct around dogs like this is: take his fears seriously. If you had gone through what he has, you’d also be scared of people.
How does this apply to clicker training, though? Clicker training is kind, non-invasive and reward-based, so should be perfect for an abused dog. Why isn’t my dog responding to the clicker?
To come up with one surprisingly common (and frequently overlooked) answer to this question, let’s take a big step back and recall a bit of the theory. (You’ll find more information about the theory in the Training Theory section of this site.)
To most dog trainers, clicker training is primarily associated with operant conditioning, in which the dog learns that a particular behaviour, like ‘sit’, will be reinforced, or rewarded, so develops a tendency to sit more often. There’s a lot more to it than that, of course, but the most widely understood aspect of clicker training is that we can use reinforcement as a technique for getting, perfecting and maintaining the behaviour we want.
But clicker training is also heavily dependent on classical conditioning, or associative learning, in which the dog learns that a particular stimulus will result in a particular outcome for the dog; after a few repetitions of the association, the dog will start responding with an anticipatory reflexive reaction to the stimulus as if it were the outcome.
It is this phenomenon which allows us to use a clicker and pair it with a food treat in the first place. We start every dog off by clicking and treating several times, until the dog shows signs of anticipating the treat. A lot of trainers refer to this as ‘charging’ the clicker, and it’s a good analogy. The sound of the clicker begins to predict the arrival of a food treat, and if the arrival of the food treat is pleasurable to the dog, the clicker becomes conditioned as a conditioned positive reinforcer, and starts triggering a pleasure response in the dog in its own right, which is why the technique is so powerful.
But that if is a very important if indeed!
Let’s go back to the rescue dog. We’ll look at the case of Orphan Annie, who has some global fear of humans and is particularly nervous about hands. Annie, whose previous owner Jemma was a real pearl because of her chronic alcoholism, has also been tormented around food. Jemma used to starve her for a couple of days so she was really hungry, then offer her a tasty tidbit and smack her or pinch her very painfully if she tried to take it, so Annie has a classic approach/avoidance conflict around food offered from someone’s hand. She wants the tidbit, but is afraid of what the hand might do.
Annie also got fed on takeaway leftovers more often than not. Jemma was particularly fond of Thai and Indian cooking, so Annie got used to sweet and spicy foods, but often had an upset stomach because of the spices, and got her mouth burned by a really hot chilli in a chicken curry once or twice.
Annie has been confiscated and placed with a new owner called Nigel, who is very kind to her and whom she adores. She has learned to enjoy Nigel petting and cuddling her, but if he moves a hand towards her suddenly, she still startles.
Nigel adores Annie too. He feels terribly sorry for her because of what she has been through, cossets her, and feeds her lots of yummy BARF food, and she is now in really good condition. (In fact, she’s a little bit overweight if anything.)
Being a responsible owner, he also decides to take her to class, and since he wants to be as kind as possible to Annie after everything she’s endured, he decides that it has to be a clicker class (and he’s not intrinsically wrong.) So off they toddle one Thursday evening.Continue reading
In 2006 a South African man, Phillip Matthysen, made sickening headlines when he used a chainsaw to decapitate his 4-month old Siberian Husky puppy after it ate his favourite parrot.
It seems both strange and appropriate that Matthysen was killed in a car accident on the morning of Sunday 12 August. He was alone in his Toyota Land Cruiser on the R50 some distance from Pretoria, and apparently rolled the vehicle. There were no other cars involved.
He died of severe head injuries.