Learning About Dogs - Teaching Dogs Article:

Teaching Dogs Vol 4 Issue 2 & 3

It is very tempting to focus your attention on teaching your dog new behaviours - in fact this is often the sexy part of clicker training, and the bit that captures us all, turns us into bores with the perpetual "Look what I've taught him now", remarks.

But when you have trained a few dogs through this, it starts to hit home that you have a LOT of behaviours "just begun" and quite a lot less actually finished.

If I were to graphically represent the time and energy that goes into finishing a behaviour I think most folk would be surprised how proportionally little of the time is actually spent on getting the behaviour initially. The definition of finished, in my interpretation, is a behaviour, robustly reinforced to the point that it will not break down, lose fluency or attitude under the assault of massive stimulus - or distractions. Additionally the behaviour should be on a minimum, effortless, easily remembered (for both me and the dog) cue - excellent stimulus control.

Cueing a finished behaviour should never leave you in doubt whether the dog will respond with less than the best it can do. It is an excellent feeling and well worth striving for.

A stimulus is anything that "stimulates" a response in the dog. The stimulus can be a sound, a movement, a scent, an emotion, an instinctive drive or all of those things at the same time. A gundog or collie will instinctively respond with certain behaviours on the scent of specialised stimuli. For the gundog a game bird for the collie the scent of sheep. These scents are their specialised predatory responses, in bred for many, many generations. In the wrong environment that same scent can be a distraction - a stimulus calling them to respond with certain specific behaviours. I can teach all my collies to sit with speed and accuracy on cue. Their experience of the ball play is the closet expression of herding behaviours and very quickly they develop the collie eye for the anticipated movement of the ball. If I hold a ball in my hand, and give the verbal cue "sit", they will very often sit slowly. The instinctive response to move with slow caution when watching prey takes a higher precedence. The reward on offer then becomes a distraction that reduces the response of the finished behaviour.

A cue is a refined stimulus that we link to specific behaviours to allow us to "stimulate" those desired behaviours. In our world the cue represents an opportunity for reward if the dog can remember and translate the cue to the matching behaviour. Many cues are not stimulus until the dog has experience of their intended link to a reward. In addition to the cold cue, perhaps "down", it can be applied contextually. A "down" cue when the dog is standing in front facing you, my require an instant drop to the poised down position, whereas in your living room, in front of the fire the response can be relaxed into a more settled position. We expect the dogs to be able to tell the difference, and being the exceptionally smart creatures they are they work it out.

A distraction is viewed as something more negative that interrupts concentration and the dog should not respond to it. Unless of course we are making the distraction ourselves and want the dog to break concentration on an undesirable response to another cue.

Concentration is really two quite different behaviours - both of which can be trained in people and dogs. A tight focus on listening for cues, translating them into behaviours, performing the behaviours as trained and secondly the concentration of duration where the youngster learns to train for longer periods without getting over tired - they concentrate for longer. If I was expecting a high concentration performance of 3 minutes I would expect training sessions: a mixture of learn-play-break, to last at least 30 minutes. This doesn't happen without building the training time over many months, and gradually you realise the dog that used to be tired after ten minutes is now raring to go after 20 minutes.

A person or dog without concentration will become distracted by the wealth of stimulus around them. The more stimulation in the environment the higher and more frequent the level of response - the ultimate "butterfly" mind that zips between responding to the varied stimulus assault.

Concentration that is the tight focus appears to eliminate responses to distractions - unrewarding stimuli. At some time in our lives we have often needed to study or work in an environment that is noisy or busy. We recognise the moment when the noise level overwhelms our ability to concentrate, which comes far sooner when we are tired. A collie (one or two from our stable) have been known to concentrate so hard on their work that they run into trees whilst working the sheep, ie they do not appear to notice life threatening distractions (and quite commonly run into their working partner).

I walked into an open plan office recently with over 50 people working, speaking to clients face to face and on the phone. When I spoke to the staff they assured me they "got used" to the noise and were able to adequately concentrate. This is the goal we need to aim for with teaching the tight focus for the dogs.

In learning terms "getting used to" a distracting noise level is learning to sort the enormous amount of stimulus into different categories:

Our brain sorts the noises, scents, sights into categories as we concentrate on the task at hand. This is why concentration in a busy environment is more tiring - the brain is multi tasking. To be able to sort each stimulus into its relevant category we need knowledge and experience of the possible outcomes of responding. Not responding to a fire alarm, a sick person, or your name can lead to problems. But the computer programmer deeply involved in their work will not respond to the chilling cup of coffee or the taste of cold coffee when they eventually take a sip - such is the intensity of their brain activities to the one exclusive task. Surprising they breath really. Computer games may seem to be useless pastimes, but they teach an inordinate amount of both tight focus skills and duration of concentration.

With the dogs, playing games will teach these skills and self reward with the challenge and continuation of the game. The closer the game is to their natural instincts the tighter the focus will become, and the games can be carefully used as cues for these foundation skills - focus (tight and of duration), motivation, attitude.

To teach the dogs to "get used to" the distractions we need to make sure that the dogs response to the distraction does not get rewarded. If the dog hears other clicks and cues they need to learn to not respond. If the dog is walking alongside a blind person they need to learn to ignore offers of social interaction from other dogs.

At the Clicker Expos during the micro shaping demonstration I use the situation to teach the dog how to dismiss stimuli as an integral part of learning of a new behaviour. I choose a dog with some experience of free shaping - from a 7 month old rough collie, to a 3 year old American Staffie. I free shape the behaviour of coming up to the steps to the podium and standing on the first step with their balance held back. This is often referred to by the guide dog trainers as "step up", where the dog must hold position on the first step of the staircase waiting for the cue to proceed from the handler. Very often when a dog moves towards the first step their intent is to continue the momentum in the direction of the staircase - as in the case of the Expo, up onto the podium (they are inherently show offs!).

With the micro shaping I initially teach the dog to hold position on that first step, and inhibit the learned response to follow through, which in this situation can be categorised as a distraction to the success of that behaviour. At the point when the dog has soundly secured this knowledge, an inhibition skill, in their mind, and is on "tight focus" I begin to introduce other distractions of a controllable level. I particularly use this moment to associate ignoring distractions whilst the behaviour is being built, since their minds seem to have the best hold on the behaviour when fluency is first achieved. Micro shaping ensures we never overload the learning expectation and reinforcement does not fall below 95%. This gives the dogs the best chance to acquire the smallest behaviours with strength and fluency.

A member of the audience is asked to stand near the dog whilst they carry out the behaviour. If the dog's response is very strong then they will be asked to step back until the dog only responds with an awareness but not an obvious behaviour. I will continue to click the learned behaviour of holding backward balance on the step up. Once I have secured five repetitions with no response to the person, I will increase the strength of the distracting stimuli very slightly - the person takes a step forward.

In all of these sessions we have been able to take a dog, with no training history with me, to continue with good quality "step up" behaviours AND a person lying on the floor alongside the steps.

The technique is to present a level of distraction stimulus that only triggers a low response, and does not offer a reward. Since the low level response of the dog gets no reward, but continuing with the established behaviour does, the dog appears to ignore the distraction. They know the distraction is present, but have filed it away into the "no need to respond" drawer.

This is often far easier to establish when the dog is on a high level of focus to achieve the behaviour, (conscious competence) rather than once the behaviour has become very familiar and slightly boring (uncomscious competence) - in that situation responding to the distraction may offer more opportunity for reward than continuing the learned behaviour. Concentration is often at its peak when "conscious competence" is achieved, but when the skill moves into "unconscious competence" then distractions are much more alluring.

Take a look at driving - concentration is fierce when acquiring the skills, but once driving becomes unconsciously competent then distractions can pop into your mind - and often take focus away from the unconscious tasks - bang! Enter tight and fast moving traffic and your driving skills will begin to tighten up with the demand on your focus. I don't wish to frighten you, but often "human errors" in aviation take place when concentration levels are low - ie boredom.

I often teach my dogs with a ball or toy that squeaks. When training, the movement for a toy, or the sound of the squeak, will trigger my dogs to join in - this is their learned response to that cue. Time must be invested in teaching the dog the experience of "not my squeak" will get no reward for responding - rather like the mother who only responds to their child's cry for attention and not all the other kids screaming as well.

You can set this up by using the "other squeak" as a cue for a game with you. If I am training alone, I put a squeaker under my foot. When it squeaks, I give the established cue to tug, and follow the success with a game. After an amazingly few repetitions the dogs will not need my cue to get the tug, but respond directly from the squeak.

I particularly use the highest level of reward for that dog. This can easily be the food. Place some treats in a bowl, and directly after the “other squeak” offer the dog the treats from the bowl. This can be transferred to a behaviour, perhaps a sit or down, after the squeak, then a click and high level reward. In time the "other squeak" will represent an opportunity for reward following some work, you can use it to trigger stronger concentrations.

The Gordons are not very turned on by tug training, but get mightily upset when they lose the rabbit down the hole. I will place a piece of food in full view of the dog - already trained to maintain focus on the food, but not help themselves, I make the squeak and inevitably they will look away, in that split second they lose concentration the food disappears. This "punishment" will only take 2 or 3 repetitions until they teach themselves to maintain strong focus on the food. Concentration skills are excellent examples of skills transferable from one situation to another. Once the concentration is held - it is easy to click and reward.

It will be worth every minute of your time to analyse a future environment and plan to teach the dog each of the stimulus that maybe become cues, and let the dog have the opportunity to file them away under "not worth responding to". Not forgetting that the novelty of a new cue is the very stimulus that pulls the attention of the dog - something new to learn, see, watch and experience.

Many dogs are over sensitive to stimulus. The types of dogs that make excellent working dogs, and are consideed easy to train, usually come with a highly tuned ability to respond to cues. There is a very thin area that is a balanced response - highly tuned enough to respond to your cues, but not so over sensitive they alert to every stimulus.

The dogs that are often "cool" about new situations, and appear not to be aroused by quantity of stimulus can often be difficult to teach stimulus control. Kent is a prime example - he is the ultimate "cool guy" at Crufts or other obnoxious stimuli circus, but is also on a constant training curve, or mountain really, of rehearsing the link between cues and behaviours. Seven sits in a row will be successful and the next one he completely forgets - I've done enough training to know my cues are consistent, but he obviously "loses" it and doesn't even seem to remember what he just did!

For dogs that can become aroused by stimulus overload we have designed a stimulus assault course.

Initially the dog is taught a range of cavaletti challenges, between the nursery level, the extended and high action movements. Four sets will be placed around the room in a square. As the dog approaches it will need to assess the height and width of the poles to analyse which movement is required. Successful completion is rewarded for each set. The dog will be able to change direction and go around the room either way, moving between all the sets without hesitation and successfully. Each dog is alone in the room whilst they train.

The team, or other participants with resting dogs, will begin to build the stimulus assault. These are broken down into different categories: sound, movement, scent, social and in each of these areas: threatening, inviting, unpredictable.

Initially all stimulus are low level, just on the threshold of triggering a response larger than an acknowledgement.

Out of sight of the dog, noise will come into the room. The dog will not be able to see what is causing the noise. Each noise will then be continued until all response has faded. By giving the dog the cavaletti tasks - both mental and physical challenges they very quickly move the stimulus to the "no need to respond" file and get on with the cavaletti tasks. The cavaletti is always off lead from free shaped learning which gives plenty of opportunity for the dog to show their comfort level in continuing.

The dog will always be given at least 30 minutes break after a session of 5 minutes stimulus assault. Through the day the dog gets to experience and file away stimulated gunshot, dinner bowls clattering, flying mats, vacuum cleaners, people dancing, sticks flying, balls tossing, doors slamming, whistles, trumpets, and any thing else our imagination can dream up.

At no time are we looking for an over response, but introducing low level versions of the stimuli to allow the dog to "file it away", and then build the intensity. The experience is very tiring for the dogs, but very valuable in building their capacity to asses, analyse, ignore and continue their tasks.