I believe to form a sound philosophy on training your dog, you have to start with a solid goal. For me, the general statement regarding training MY dog would be to have a well behaved, home and hunting companion that I can count on to bag the most birds. That being said, my ideal dog would do the following when hunting: Cover as much ground as possible within gun-range, have a good enough nose to not over-run birds, hold point until I can set-up shooters for the most ideal shot, hold until released, retrieve the bird to hand (if I hit it), and repeat the above until I have had enough (i.e., am exhausted, have a full limit, have a sore shoulder from shooting, etc...). I have not worked my dogs in field trials, although I have been interested in FT as a way to keep the dog in working shape during the off-season. Once this high-level goal is in place, the next philosophical conversation is about how to get there (train the dog).
It is incredible to me how many philosophies there are on training dogs. I have read so much material on different philosophies, and slams on other philosophies of dog training, it was enough to make my head spin. Use of a shock collar, ban of shock collars, citronella spray, shaker-cans... As an exmaple, when my parents had Buck (our first GSP as a family when I was in High School), we used about every holistic, and weird contraption to make the dog slow-down in the field and listen. There was this chain-thingy that you would shake with a tuning fork in it... Balls clipped to his collar that would bounce between his paws and head... Terrible torture devices! Turns out that the dog went deaf after we got him (he was about 7 when this happened). Poor thing didn't know we were recalling him because he couldn't hear!
Then there was Heidi (our family second GSP). We put her through obedience courses (R+, more on that later), but when it came to the field nothing could slow her down. She constantly ranged, and busted birds way out of range. We could not get her to accept a reward (R) in the field - she would spit it out.
My overall philosophy began to come together with MY first dog, Sophie. A friend of mine (who hunted ducks, and worked a golden lab) talked with me, and introduced me to P- training as a means to condition the dog to do what it is supposed to, and understand P+. Now, I had a better understanding of what is referred to as "Basic Learning Theory", or "Behavioral Management" (actually the "Behavioral Management" concepts related to a management course I was taking at the time).
There is actually a good blog on this website, along with a link to a first chapter of a book that describes the concepts pretty well: www.shorthairs.net/GSPForum/tabid/220/forumid/2/postid/88038/view/topic/Default.aspx. In the chapter, the authors (Jim Barry, Mary Emmen, and Susan Smith) describe the operant training in terms of the four quadrants (R+, R-, P+ and P-, see page 8 in the link: www.clickertraining.com/files/PosGunDog_Chapter_One.pdf).
So, to me, most effective training of my dog uses all 4 quadrants of the operants - because, why would I want to limit myself. For example, R+ Training (taken to an extreme) only uses 25% of the available operants, so why would I expect R+ training to be more than 25% effective? Of course, this is an oversimplification. I also believe that effective training is a function of the intensity of the operant, and time applied. And what I mean here by intensity is not "the more intense, the shorter the time", but rather the correct operant used with "OPTIMAL" intensity, the shorter the time. I believe optimal intensity is a function of the comfort of the trainer (condition of the trainer (comfort level) can be an operant in-and-of-itself), and the animal being trained. It is also important to state that it is possible to apply TOO MUCH operant, as well as TOO LITTLE. Because of this, the optimal intensity range of operants may be defined in a green band within the operant chart, assuming the axis represents intensity. Application of an operant too far outside of this theoretical "band", and the time to train increases. I also believe this describes the challenge faced by trainers - to select the appropriate opperant in the appropriate intensity. This is also the reason I choose to train my own dog - after all, who would know my dog better than I. Then again, a good trainer would be removed enough from the dog to apply higher amounts of operant (particularly Pressure (P)), and be aware enough to not apply too much. Enter the second factor behind training my own dog - Budget - but I will not discuss that here.
So any concept can then be broken-down into individual parts that can be reinforced with appropriate opperant as follows (in my mind):
- Demonstration - make the dog do what you want, and Reinforce (R+)
- Teach the dog to make the right decision (follow command that was demonstrated) by applying pressure, and removing when compliance is achieved (P-). This also prepares the dog for the next step where P+ will be used specifically by demonstrating, and teaching the dog what removes the pressure (i.e., compliance).
- Reinforcement - dog should be complying without pressure applied, if the dog does not comply, Pressure is applied (P+). Here also, I like to throw in a little R+ for compliance, as well as R- (ignore behavior) if performance does not meet my expectations. An example of R- for me would be a recall (Come, or Whistle), or a fetch where the dog does not position itself properly - I would ignore until she is in the correct position, then R+ (if appropriate based on history).
I have found this approach works well, regardless of the command being taught - Sit, Heel, Kennel, Come, Fetch (yes, Force-fetch follows this as well), Hunt (although the dog usually does this eagerly anyways once it is demonstrated, so I don't directly follow it, but R+ when appropriate Range is achieved, or for a check-in by the dog, and enforce Come, or Heel when it is not).
Finally, I believe in consistent repetition in various environments to enforce habitiation. I think that (probably because of my statistics and quality engineering background) by repeating obedience that my enforcement tends to the average (hopefully in the interest of the dog), and mistakes made are smoothed-over in subsequent repetitions. In other words, I admit that I am human, and I admit that application of the operant at an "Ideal" intensity is probably not feasibile (also because of factors with the dog - they have bad days too). Repetition means that I have more opportunities to correct my handling, and the response of the dog when I do make a mistake.
Dogs are amazing animals, and mistakes by handlers, if not left to systemic habits, will be forgiven, and corrected. The important thing is constantly evaluating the dogs behaviors, and actions, as well as what has potentially caused them from a training perspective - as a responsible handler. The key virtues that I feel I must constantly hone as a trainer for my dog are persistence, and patience. I have not found any behavioral issue with my dogs that a combination of these virtues will not fix.
So there it is, my general comments on my philosophy of training my dogs.