Concept to Context
Professor Tom Brownlee and Certified Trainer Molly Rowland
Virtually since the time of Max Von Stephanitz we have known that dogs are context specific learners. We mold most of our training around this fact, realize it or not. A second factor that we assimilate so well after training a few dogs that we often forget to recognize it is that all training is incremental. In the early stages of dope training the dog is worked on a line of crates, in my case three, then a line of boxes. Basically, this introduces two increments at approximately the same time. One is a transition from being able to see and smell the target/reward through the crate to relying solely on olfaction to find the target/reward in the boxes. Two it starts the dog on recognizing the context of a line search. This context is then expanded upon by adding more boxes etc. to the line. (Another increment...) Then expanded further by mixing up boxes with other forms of containers. You will notice at this point the dog no longer cares what kind of containers are in the lineup. He now has the context. He does not know boxes from backpacks from Shinola. He just recognizes the line and will search it regardless of its component parts. He has now gone from 'context' to 'concept.' Your next increment is, of course, to train the line search in a number of various venues, again gradually introducing various distractions in an incremental manner so the dog learns to work past them. Thankfully the high drive dogs we usually work with get past them easily if the reward is high enough.
Another example, albeit a bit more complicated would be vehicle searches. Without going into all the increments involved in this training we will go straight to the context. Like any other training scenarios and venues have to be introduced to cover the spectrum of possible future searches. Say you start with a sub-compact car. Will the dog instantly know what to do when presented with a tractor/trailer truck to search? Chances are that he will not. One of the oldest adages in dog training is that “ When you try something just to see what happens...nothing does.” Again, you train each type and size of vehicle incrementally until the dog recognizes all the different contexts he will be searching. In this case different types of vehicles. Soon enough they will all be grist for his mill and a simple search command will put him in the mode for any vehicle. He has now generalized the 'concept'.
Dogs are capable of conceptual learning on an elementary level, and know it or not you take advantage of it every day. The dog's day and his training are full of what trainers call “triggers”. Dogs, especially the breeds we work with are often considered by society as being noble altruistic creatures. To some extent, this is true. The bottom line, however, is that they are very much like us in that they are self-serving creatures interested in their own gratification. By domesticating them we have taken care of their most basic needs, food and shelter. The only thing left for them to concentrate on is the hunt and the kill which we have morphed into a game. Their bottom line in all this is how to get to that kill (reward) by the quickest means possible. It is your job to help them do that by training shortcuts to the reward. More incremental training that enables the dog to process a search or a track quickly to the benefit of both of you. To this end, some of this conceptual thinking on the dog's part kicks in. On any given day the dog is exposed to what we call triggers. Little events that signal bigger events to come. The dogs watch what you do to a degree that is staggering if you'll notice. Putting on your uniform is a trigger. Dog knows he's going to work. Putting on your flip flops and shorts is a trigger. Not going to work. Stepping on the gas in your patrol car, turning on the light bar, hooking a leash to the collar, hooking lead to tracking harness, etc. everything triggers the dog for what’s coming next and if you'll notice he learned them all by himself...because they lead to a reward. My narcotics dog only needed to see the Kong to know she was going to search for dope, and it didn't matter where, vehicles, buildings or open air. These are all concepts in their simplest forms.
More complicated examples come to mind for all of us, but the first in my mind came from training Mobility Assistance dogs for the disabled. They must be trained to open doors. Hence, you train the dog to push a French door handle like a lever. You train the dog to push 4-inch square Handicap Buttons, or 8 inch round ones, or worse yet the little red ones in blue boxes. The dog has to learn how to work all of these, and you know you've “got him” when you can walk him up to any door and ask him to “open the door” and you see his eyes searching for what type of button or handle the door has. He has gone from the 'context' of different types of buttons and handles to the 'concept' of “open the door” Law Enforcement K9 is full of similar examples, and I'm sure you can come up with a few at a moments notice
In the course of events while being exposed to numerous dogs we can't help but notice some are better at grasping conceptual learning than others, and of course, we consider that grasp a good thing in a working dog. Questions arise. Why the difference? And what can we do to enhance the transitional process between context and concept?
A little analytical thought and the answers become readily apparent. First is the inherent intelligence of the dog. This has never been quantified scientifically. Pseudo-scientists have tried to compile a high to low dog IQ list....but their “research” relies on much the same premise P.T. Barnum relied on. We can only generalize, and like others, insert our own bias. The herding breeds we work with seem to have a decided advantage in the intelligence department. ( Yes, I am very biased...) The Labradors we often work with on narcotics, at least in my experience, train on the same timeline as the herding dogs.
So why would this be? The answer is simple...its drive. The herding dogs, and the labs were bred to work, bred to have the energy and the musculature to do it. Along with this comes a craving for both physical and mental stimulation. This craving for mental stimulation seems to equate to intelligence, but why?
If you have attended any advanced schools taught by Dr. Aycock S.M.I. You have repeatedly heard about the value of the reward. This is where we begin to see how we can accelerate the transition from context-specific learning to conceptual thought. If the value of the reward is high enough the dog will naturally try to seek out any shortcuts to its acquisition. It is up to you to train these short cuts. Witness for example an experienced narcotics dog when he enters a vehicle for a search. If he doesn't hit the scent stream immediately he goes into a process of elimination, by checking high, low, and in between. He does this by checking places he has previously found narcotics and quickly eliminating them, so as to funnel himself towards the goal as quickly as possible. Like many of you others, I try to accelerate the process. If the dog is in the vehicle for more than 30 seconds without hitting the scent stream I change the air flow, which usually helps him get on it, and get to his reward. The same can be said for building searches for suspects, or for re-acquiring a track after a negative. This all stems back to the value of the reward. It needs to be Big-Time for the dog, so he will actively seek these short cuts to obtaining it. Every dog is slightly different and each will let you know what reward they value the most. Hint: it is usually one with you attached to it, being 'interactive'. The tennis ball is more fun if you are throwing it, Kong on a rope is more fun if you are on the other end of the rope. The toy/reward is ten times more appealing if you are attached to it. It is also why many dogs consider the bite to be the ultimate reward. They get the hunt and the “kill” on the biggest liveliest tug toy ever!
Enter the BIG variable to get the dog to conceptual thought...your training. Positive reinforcement in training any of the disciplines we do is critical to success. If the dog doesn't catch on, or “get it” right away being frustrated, punitive or compulsive will do nothing to accomplish your goals, and indeed will set them back and in some cases shut the dog down and make them unobtainable. The best quick example is tracking. Once a dog overshoots a 90-degree turn in the track, the idea is not to yank him off his feet and scream at him in an effort to issue a 'correction'. Simply stop him and start 'modeling' as if you were looking for the track yourself. Throw in a little verbal encouragement “ check over here
Bud...” The idea, whether its tracking, dope work, or bite work is to HELP the dog find the shortcut he is looking for towards his reward. Don't give him the devil for not figuring it out, help him through it!
Again it is imperative to vary things up in their context-specific learning. Train the dog in different types and sizes of vehicles so that he recognizes the context of say, a truck. He will see it and think “ I got this....” We trained one dog on vehicles, but the day came when we had to search a piece of construction machinery as I had never seen. The dog had enough experience in everything else that he did not hesitate to go into his search pattern.
A few short years ago SMI Chris Aycock contacted me after attending a series of Neuroscience Lectures by Dr. Robert Sapolsky, excitedly exclaiming “ This explains EVERYTHING!” As you know, Chris went on to get a PhD. in Neuroscience and has gone on to hand as much pertinent information as possible down to the rest of us. There is indeed an explanation for some of this to be found in neuroscience. From my protégé 'Molly Rowland ( remember that name....you'll hear it again.) and current Advanced Neuroscience student of Chris' comes this...“ A neuro principle called 'pattern completion'- The activation of an entire pattern by a subset of the original experience. When we encounter a subset or portion of the original experience that originally established the memory trace it can activate or replay the entire experience. This implies that there is some related piece of an event that may allow a dogs brain to activate a pathway to behave correctly even though he hasn't experienced that event in particular. i.e. – you've tracked a field before but you haven't tracked this particular field before... “ She concludes with” The hippocampus index representation binds components of one episode into a representation of the context. Thus, encountering some component of the episode activates the index that in turn projects back to the neocortex and activates the cortical representation of the context.” In short, allowing the dog to morph 'context' into 'concept'....
An important adjunct to your training is your communication with the dog. Again, frustration is counter-productive. You are teaching him a language, which is foreign to him English. Repeating it and raising your voice doesn't make the dog understand it any better and indeed is usually counterproductive. Watch and listen to your best handlers-in any training discipline- and you will notice they rarely if ever raise their voice above-normal volume. Once in awhile raising your voice acts as a motivator, when the dog is on a bite or in pursuit. Cheering him on lets him know he is doing the right thing. Never forget that communication is a two-way street. Dog’s first line of communication is not verbal, its body language. It is imperative that you be able to read and understand it. You can tell by the angle of the dog's ears, by his gaze, by his posture, etc. whether or not your training is getting through to him. If you can read his signals, you can tell when you are succeeding, or when you have hit a wall. As I have told every student I've ever had, if you don't learn more from your dog than you learn from me, then you are doing it wrong. Without exception, those dogs with an exceptional skill set and good conceptual thought have handlers who communicate well with them. Remember that all training is incremental and if your dog stumbles on an increment, it is up to you to drop back, think, and figure out a way you can help him get through it.