There is a section about SAR work with dogs on the bottom of this linked page (Dees Dogs/Nancy Lyon). It has a great description of developing canine partners who are persistent problem solvers. It also shows a great appreciation of the partnership, with full respect for the abilities of both parties.
Working dogs on stock has similarities to SAR work in that obedience of the dog cannot come at the price of initiative. Training dogs for these jobs is as much about developing the dog's natural instincts, problem solving skills, and most importantly the dog’s confidence in their own ability to handle the situation on their own as it is about training the dog to accept our leadership. In the end a tremendously obedient dog with limited skills and/or no confidence to apply those skills without constant direction is not a valuable partner. A dog with tremendous skills and confidence in the chosen job but of a less cooperative mind is indeed more valuable though likely frustrating to handle.
Early training on stock is all about developing a dog’s abilities. My early dogs were quite confident that they could do the job and eager to take control, certainly no lack of willingness to take initiative. I was able to focus on commands while the dogs capitalized on every opportunity to work out problems for themselves. These were dogs who naturally focused on the objectives, interpreting my commands as information about where I wanted the stock as much as instructions to execute a specific move themselves. Since then I’ve had some more sensitive dogs that have been wary of making mistakes, wanting explicit instructions, lacking the overriding initiative to control the stock. One of these dogs, my Song, was eager to take responsibility, but any correction left her worried about being wrong and reluctant to try again. Since her behavior was not egregious and the stock was not overly stressed I allowed her a large amount of freedom in early training. Eventually she matured to a point that she could recognize corrections as information from me to try another method. From that point she trained up extremely quickly. Fina, who was bounced around a bit for training and handling when young, is extremely sensitive. I have had to work harder to get Fina to take initiative and do what needs to be done to control the stock. She worries about being wrong, doubts her ability to solve problems on her own, and thinks of commands as simple instructions as to her next move. She does not naturally think in terms of objectives, where I want the stock. I’ve spent some time this winter doing farm work with her, denying her specific commands, and holding her responsible for the job. We had a few uncomfortable sessions, with her wanting to wait for specific commands and instead getting a sharp bark of her name when sheep took advantage of her. She struggled, unhappy with the lack of instruction, but she kept thinking. Then after a few rounds she began to understand. She realized she was responsible for putting the sheep somewhere and keeping them there while my back was turned and I was doing something else. She realized I had little interest in her specific moves, as long as the job got done. She got the chance to work the moves out on her own and SHE LIKED IT. This newfound sense of objectives and responsibility is fragile as she is a sensitive dog. I’ll need to be careful to foster her new abilities while reminding her that she needs to work to my objectives.
How much responsibility rests with the dog? That depends on the handler. Some folks like a dog that is infinitely and immediately biddable. I prefer a dog that will accept my objectives, but be willing to disagree with me on the method. I’ll accept non-compliance if I feel they are still working to get the job done. If I know I’m giving a command that the dog will find unwise I’ll insist by using a stronger voice or whistle. In this way I tell the dog that I am overriding their responsibility for this move.
The balance between obedience and responsibility for the stock takes work to maintain. The more we tune our dogs for the immediate compliance required in trials the more we eat away at their initiative in controlling the stock on their own. Time off from trialing with some farm work helps to keep the dogs focused on the objectives. My Cato dog was easy in this respect as no matter how much we trained it would be a cold day in hell before he relinquished control of his stock. Yet he was from the beginning willing to partner with me and allow me to define the objectives.