|
Dominance(?!?), Fighting, Biting, Compliance Constraints and Learning Raising the Bar in Dog Training Presented by: Ian Dunbar Friday: Dominance (?!?), Fighting, Biting, Compliance & Punishment"Dominance" is most certainly the most misunderstood topic in dog behavior and training. A misunderstanding of wolf behavior has been applied to dogs, and then a misunderstanding of dog-dog behavior has been cavalierly extrapolated to dog-human interactions and training. Misunderstood notions of supposed physical dominance-hierarchies are often used as rationale (an excuse) for physical and painful punishment when training dogs and especially when attempting to resolve biting and fighting problems. These simplistic views of dog behavior are an insult to dogs. Domestic dogs living together develop extremely complex and sophisticated social structures and dogs living alone require impeccable social savvy to meet, greet and get along with every unfamiliar dog that they meet. (Human social skills pale in comparison.) More disturbing though, the unwarranted application of physical dominance in training has had an extreme negative impact on the quality of life of dogs and their owners. The mere use of the term "dominance" biases the training arena, encouraging an adversarial relationship between people and dogs and prompting people to attempt to physically dominate their dogs. Physical domination causes resentment, mistrust and fear, and ultimately creates yet more problems. The dominance perspective is just so wrong. And the mere suggestion of physically dominating dogs is downright dangerous in pet dog training in which a good third of our clients are children. Most so-called "dominant" behaviors, such as jumping-up, pulling on-leash, housesoiling, humping, eating first and going through doors first, may be quickly and easily resolved by mind-boggingly simple training techniques, such as all-or-none reward training. Certainly some dogs, especially males, have pushy personalities with people - always testing the rules, however, instigating more rules and calmly insisting on improved compliance usually puts the dog back on track. Similarly, some dogs tend to "bully" others and require immediate, continuous and representative feedback regarding their appropriate and inappropriate play styles. Even so, most categories of aggression, especially including inter-male, dominant, territorial, maternal, and idiosyncratic aggression, etc., appear to be fear-based, as diagnosed by successful treatment primarily by classical conditioning and progressive desensitization. Topics include:
Saturday: Constraints on LearningLack of compliance, or more specifically, low response-reliability, is due in part by selective attention and canine sensory constraints on signal input. However, most learning is constrained via the nearly impossible task of attempting to correctly apply a rigorous yet extremely cumbersome learning theory to real-life events. Existing learning theory is long overdue for a radical overhaul. Dog trainers are in dire need of a Learning Theory Redux that is efficient and effective in practice. Learning Theory is approximately 110 years old. Much has held up over time (especially in laboratory settings) but much is outdated and simply doesn't work that well in real life. The extreme effectiveness of consistently applied punishment in laboratory experiments was a major reason for the increased use of punishment in animal training throughout the 20th Century and now, the many constraints of learning theory in practice are a major reason we are currently seeing a resurgence of the use of physical punishment. Learning Theory was whelped from dog training and nourished by thousands and thousands of laboratory experiments, whereby consistent computers trained animals (manly rats and pigeons) using quantum rewards (food pellets) and punishments (electric shocks). However, people are not computers and do not have the ability to work tirelessly and consistently for hours on end. When people train animals, many theoretical rules fail in practice, imposing severe constraints on learning. Most reinforcement schedules and punishment schedules are impractical and ineffective and the misuse of punishment causes a variety of additional problems. And when reward-based results fail to reach par, dog owners seek alternative methods. On the other hand, being human (rather than a computer) has huge advantages when training dogs. The power of relationship and the use of language and voice (analogue and instructive) allow people to deliver much more sophisticated and effective feedback than any computer. Language enables people to avoid the many constraints of laboratory learning theory and to absolutely transcend the training ability of any computer (with its precise and clinical, yet non-instructive, quantum feedback) . Dog trainers desperately need to completely reevaluate current learning theory and devise a simpler, more effective "applied learning theory" that is much more easily understood by pet owners. We need simple, precise and unambiguous terminology, basic learning sequences, instructive and analogue binary feedback, (especially calm, gentle and insistent instructive reprimands), and reinforcement and punishment schedules that work in practice. Topics include:
Sunday: Raising the Bar in Dog TrainingCompetition dog training comprises specialized dogs, experienced handlers and a finite curriculum with examination questions known well beforehand. Pet dog training comprises an unknown, infinite curriculum, largely inexperienced handlers and dogs of all types. The divergence of pet dog training from competition training has caused many changes. By and large, training techniques have become much more user-friendly and dog-friendly (with oodles of classical conditioning, food and toy lures and rewards, and lots of fun and games), but... criteria, speed of acquisition, precision and ultimate response-reliability have all taken a nose-dive. What happened to emergency sits and downs, rock-solid stays, off-leash control, snazzy heeling and calm on-leash walking? With the absence of periodic quantification the success of pet dog training has gone downhill. And when training doesn't work well, dog owners either blame the dog or seek help elsewhere. So many owners respond to noncompliance by punishing the dog even though it can be proved that most dogs fail to comply simply because they have not been trained sufficiently; the dogs do not fully understand the meaning of instructions and they usually have not been motivated to respond. The goal of training is to produce reliable, distal, verbal control without the reliance on any training tools. Knowing what your dog understands by your commands is important so that you do not become frustrated (and blame the dog) for poor performance. Instead, poor responses from the dog should prompt owners to re-evaluate their training techniques. Quantification is proof-positive of your dog's performance-reliability and level of command-comprehension, (or lack thereof). Quantification allows you to set realistic, personal long-term training goals. Quantification offers irrefutable proof when you reach these goals (and deserve to congratulate yourself). Basically, if you keep score, your dog’s performance will improve. Quantify for Quality! However, even once dogs have a solid comprehension of training commands, that does not mean that they will necessarily do what we want. On the contrary, reliable and precise performance is all about motivating dogs to want to do what we want them to do. Once the dog is on our team, then only very occasionally is it necessary to enforce compliance, without of course, the use of force. Topics include:
|