Fast Track Dog Training Part 2
Over the last 100 years or so scientists have been investing much time in studying animal behaviour and how learning occurs: often, so as to better understand the working of the human animal. More recently great interest has developed in dogs per se. Dogs are big business these days with dog training programmes on television becoming both popular and controversial. Many universities now schedule courses covering animal cognition and learning theory and dog training is moving from being a skill that the local ‘dogman’ has a knack for, to an academic subject. That said the academic courses are not able to teach intuition - the natural ability with which some people are born.
Now is an exciting time to be involved in dog training: a time of enlightenment, a time of great progress. We are fortunate to be involved in dog training in general and especially working trials at this time. We can take the information the scientists have gleaned from their experiments and utilise it to hone our training methods.
In 1959 Lubow and Moore carried out experiments with sheep and goats and discovered the phenomenon they termed ‘latent inhibition.’ The use of knowledge gleaned from these experiments can have a significant effect on the time taken to teach our dogs an exercise; whether to ensure the dog learns something new as quickly as possible, or to help prevent the dog learning things we do not want it to learn. Both very useful for our competition dogs.
In the experiments with sheep and goats Lubow and Moore exposed one group of animals to a flashing light. After ten repetitions of seeing the flashing light with no consequence (nothing happened), the flashing light was then paired with a shock.
Another group of animals (the control group) were given the same flashing light paired with the shock. However they were not shown the flashing light before it was paired with the shock. They quickly became fearful when the light flashed.
Animals in the first group (the pre-exposure group) took much longer to learn that the flashing light meant a shock was coming. They had learned to ignore the light – it had no significance.
For practical example of this in every day terms imagine that every morning when I put my kitchen light on it flashes two or three times before coming on. Then one morning I put the light on as usual, but the kettle does not work. I phone the electrician and tell him that my kettle will not work. I have no other information – there was no indication of a problem and everything else is fine.
Now for comparison, imagine that when I switch my light on in the mornings it comes on straight away, it always does. Now imagine that one morning I go into the kitchen and switch on the light and it flashes two or three times before coming on, then I go to boil the kettle and it does not work. I immediately phone the electrician and tell him that the light flashed and the kettle does not work. I have associated a connection between the two events. In the first example I had learned to ignore the flashing light (just like the sheep and goats in Lubow and Moore’s experiments) – it had no significance.
Okay so what have flashing lights to do with dog training? Suppose I were to tell you that a similar response was elicited from a sound. That the animals who heard a tone numerous times before it was paired with a shock took longer to make the connection between the sound and the shock than animal who did not hear the sound before it was paired with the shock. Forget that the experiment involved a shock. That was used because it is a good way to demonstrate if learning has occurred. The point is that stimuli (a noise or visual stimulus) that have no apparent predictive value come to be ignored. If a dog repeatedly hears a noise (or a command) that has no outcome for that dog then it learns to ignore that noise, or is slower to connect that noise with an outcome at a later date.
An obvious first example of using this information in training the competition dog would be a dog that hears a steward (or instructor) say ‘recall your dog’ (e.g. in a recall or back over the scale) and the handler calls his dog. Very quickly the dog learns to recall on the steward’s command and marks are lost for anticipation. However if we used our knowledge of latent inhibition and had our helpers shouting "recall your dog" numerous times in our dog’s early training when we had no intention of recalling – when we were just teaching the dog to sit and wait for example, the dog would ignore the stewards command because he has learnt it means nothing because of his pre-exposure that meant nothing. In this case it would be much easier to prevent the dog responding to the stewards’ commands throughout its working career. The handler should also remember not to always respond immediately to commands shouted by helpers and instructors in training to prevent the dog learning the unwanted association.
That is one way we can use the theory to help prevent the dog learning things we do not want it to learn. I am sure you will find many more uses. I find it very helpful to teach a young dog to ignore poles when tracking so that when he heads towards a freshly prepared search square whilst still tracking at a trial he will not be distracted from tracking – he will not just pull me to the pole! Bearing in mind that I teach a pole association for the search square and I teach my dogs to run to a toy on a pole in early sendaway training it would be easy for an inexperienced dog to be drawn towards a pole when tracking. Using knowledge of latent inhibition we can prevent this happening.
In the very early track training I leave poles dotted around the field, poles that have no use or meaning. They should be placed before training begins (before you get the dog out of the car) making sure you do not foul the tracking land whilst placing the poles. See Fig 1.When the pup comes out to track he quickly learns that poles (apart from the start pole) have no significance when tracking. If you can leave poles out in the field so there are no tracks to them great, you can just lay your track straight through the four poles. Otherwise you need to walk further to avoid laying cross tracks.
Later when teaching the dog to track against differing wind direction latent inhibition can again be used to prevent the unwanted pole association.
Practicing starts from poles (as illustrated) can also be beneficial. The dog should know that eyes are not needed for tracking!
After this type of early training the dog will be less likely to simply pull you towards a pole that he can see, even when in UD the pole indicates the direction of the first leg: he will use his nose to track. We do not want him to simply pull towards a pole. When the tracking harness is on, poles (apart from the start pole from which you indicate the track) are of no consequence.
Now you are able to use latent inhibition to avoid unwanted associations it is easy to understand how to speed up training by not exposing the dog to stimuli (e.g. commands) that later, you will want him to learn.
For example, in early ‘speak’ training, if you say "speak" ten times before your dog eventually barks he will be unlikely to associate the command with the action when he does bark. He will be learning the word has no significance - you keep repeating it when nothing is happening.
You may have heard clicker trainers say ‘get the behaviour before you put it on cue.’ It is for this reason -to prevent the dog learning the command has no significance.
If you say ‘speak’ and tease the dog with a ball to make him speak it may take many repetitions before the dog barks. That means he has ignored the word many times – it meant nothing. Why not simply tease the dog until he barks and reward him. Repeat this many times until you know you can make him bark by just teasing him and rewarding him, only then prefix the teasing with the command to speak. Hey presto a command with a meaning!