Dog training: The case for using food lures and rewards
Monday August 3rd, 2009
Lure/reward training is efficient
Lure/reward methods take less time than trying to train using corrections and punishments. By nature, lure/reward training is much more efficient than punishment-training. Whereas there are an infinite number ways for the dog to get it wrong (which require an infinite number of punishments), there is only one right way! So in terms of your time-investment and your dog's speed of learning, it is far better to show your dog exactly what is required and to reward it for complying, than it is to attempt the impossible-trying to punish the dog for each and every mistake.
Housetraining is a fine example. The average dog could think of an infinite number of choice locations to empty its bladder and bowels in the space of a 200-square-foot living room alone. Since the dog could make an infinite number of mistakes, correcting the problem would require an infinite number of corrections, and hence an infinite amount of time. Punishment-training is like the Myth of Sisyphus-an everlastingly laborious and theoretically impossible task.
On the other hand, I can think only of one appropriate place for my dogs to relieve themselves-in the P Zone (underneath the plum tree at the bottom of the garden). Consequently, using lure/reward methods, housetraining becomes a quick and easy process. All we have to do is show our dog where we would like it to relieve itself, and then praise and reward it for doing so.
The same principle applies to other potential behavior problems like chewing, digging, and barking. Teach your dog what to chew, where to dig, and when to bark, and then reward it for doing so. We may extend the same principle to teaching manners. Rather than punishing a dog for jumping up, teach it to sit when greeting people and to jump up only when requested to "Give a Hug!"
There is an additional reason that punishment-training takes more time: the trainer has to wait for the dog to misbehave before it can be punished. Now if that isn't just too silly for words, I don't know what is. Aside from being decidedly unfair to provide the dog with no education and then punish it for breaking rules it never knew existed, what a ridiculous delay in training-to fiddle about and wait around for transgressions from an untrained dog. Why not be proactive and just teach your dog how you would like it to behave? Lure/reward training allows you to take the initiative and teach your dog what you would like it to do before it is forced to improvise and make mistakes. In fact, with correct management, lure/reward housetraining and chewtoy training can be virtually errorless.
Lure/reward training is effective
Dog training techniques must work: that is, they must be effective. Otherwise training would be a waste of time.
Punishment-training is relatively ineffective for pet dog training. People are just too inconsistent for punishment-training techniques to work in the domestic setting.
Many dog books emphasize consistency as the hallmark of successful training. Consistency is essential if one attempts to enforce household rules and manners using correction and punishment. For punishment-training to be effective, the dog must be punished each and every time it misbehaves. Now, apart from being a lot of work, this is actually impossible in most households. People are simply not 100% consistent 100% of the time. Certainly people can concentrate for short periods, but not all the time. Even when people try their hardest to concentrate for limited periods, their attention often wanders.
The myth of successful punishment-training comes from scientific research. Animal learning theory evolved from literally thousands of experiments involving millions of laboratory rats and pigeons, many of which were effectively trained using punishment-training (plain and simple punishment, aversive conditioning, and avoidance learning). In the laboratory experiments, the animals' behavior was monitored by electronic and mechanical sensors and punishment (usually electric shock) was automatic, or administered by computer.
Although punishment-training works extremely well in the laboratory, where animals are trained by tirelessly consistent computers, punishment-training is notoriously ineffective when people train animals, or when people teach people. In fact, were unfortunate dogs not abused in the process, punishment-training would be one huge joke.
Aside from being a colossal effort, and not working particularly well in practice, punishment-training has yet another major drawback. The dog only has to misbehave without being punished once to learn that there are occasions when it will not be punished for doing what it likes. This creates a multitude of problems, including owner-absent behavior problems, owner-present but functionally-absent problems, and owner physically-present but mentally-absent problems.
- Owner-Absent Problems: The dog learns it would be a mite foolhardy to act like a dog when its owner is present, and so it intelligently waits for its owner to leave before indulging its basic doggy nature. The dog learns to enjoy expressing its normal, natural, and necessary doggy behavior (usually in a manner which owners consider to be inappropriate and annoying) when the owner is physically absent (out of the room, or away from home). Thus, punishment is often a primary motivator for owner-absent housesoiling, chewing, digging, and barking. In a sense, the so-called "treatment" is the cause.
- Owner-Present but Functionally-Absent Problems: The dog learns it cannot be punished for misbehaving when it is off-leash and out of reach, or when the owner cannot respond (chatting on the telephone, taking a shower, cooking at the stove, feeding the baby, talking to someone on the street, or driving a car). To make matters worse, these are all extremely inconvenient times for your dog to act up and misbehave.
- Owner Physically-Present but Mentally-Absent Problems Dogs quickly learn to discern those times when their owners are mentally absent, and not paying attention. The less said about this the better.
But now the good news! Whereas consistency is absolutely essential for punishment-training to be effective, there is no need to be consistent when lure/reward training. In fact, inconsistency can actually be advantageous when rewarding a dog. Whenever your dog complies with your wishes, you may reward it if, and when, you like. Distribute the rewards whenever the fancy takes you-totally at random, if you like. You don't have to reward your dog every time. Isn't that just wickedly wunderbar?
Consider, for example, the allure of a one-armed bandit (slot machine), which dispenses a variety of rewards at random, compared with the dull predictability of an ATM, or a food vending machine, which pays out all the time (or at least is meant to). It may seem strange, but we humans will actually work harder and longer for fewer rewards if the rewards are unpredictable. Nothing spoils a dog more quickly, or devalues rewards in training, than handing out rewards willy nilly for every remotely correct response. What's more, when the dog is rewarded all the time, it takes only one response without reward for the critter to go on strike. One failed attempt to get food out of a vending machine and we assume it's broken. We stop trying. Similarly, one unrewarded "Sit" and your dog surmises, "Ahhhh Hah! I don't think she has a reward to give me," and the dog stops trying.
So adopt the slot machine approach. Make the handouts appear to be unpredictable, so your dog learns it is not uncommon for a number of unrewarded trials to be followed by a big payout. Vary the frequency, type, and amount of praise and rewards and you'll find your dog will gladly oblige without having to be rewarded each time. This puts you in the driver's seat. You have discovered a powerful and enjoyable means of motivating your dog and modifying its behavior to your liking.
Lure/reward training is easy
You know, saying "Whattt a good doggie," patting, stroking, scratching the little critter behind its ear, or even giving a food treat is all so effortless, while reprimanding or punishing a dog often requires considerable effort. One has to get up, get stern, and get over to the poor dog to deliver the punishment. Moreover, most punishment routines involve physical manhandling. Pushing, pulling, shaking, or "alpha-rolling" the dog can be quite tiring, and well beyond the physical capabilities of most novice dog owners. Considering that many dog-owning families include children (who try to mimic their parents' behavior), attempting to control a dog by physical punishment or force is not only cruel and stupid, it is potentially dangerous.
Dog training techniques must be effective, but they also must produce the desired effect without unwanted side-effects. Otherwise training could be counterproductive.
When reward-training goes awry, your dog may not completely master what you want it to learn, but it does learn to enjoy your ineffectual attempts at reward-training, and it does learn to develop a fondness for its hapless trainer-the reward-giver, i.e., you.
On the other hand, when punishment-training goes awry (as is often the case), your dog does not learn what you want it to learn. However, it does learn to dislike training, and to dislike its trainer, i.e., you. Punishment usually creates more problems than it resolves.
By definition, punishment decreases the frequency of immediately preceding behaviors. Apart from being de-motivating for the dog, repeated, ineffective punishment quickly erodes the very foundations of the relationship between dog and trainer. Also, technically, since ineffective punishment does not produce the desired effect of reducing unwanted behavior, the inflicted "nasties" may not accurately be defined as punishment. If punishment does not decrease unwanted behavior, then it is not punishment. It is merely harassment. Abuse-pure and simple abuse!
Lure/reward training is enjoyable
Now, call me a worm if you like, but instead of having to play the bad guy, I would much rather praise my dog, pet her, and occasionally offer a tasty treat. And, as it happens, my dogs have all informed me that they much prefer my grateful, affectionate, and generous teaching-mode to any outdated, adversarial, and authoritarian approach. This means we are all happy.
So what have we got? User-friendly and dog-friendly dog training. I mean, if reading novels, watching television, eating chocolate, or playing golf were not enjoyable, we wouldn't engage in these behaviors, right? Similarly, if training is not utterly enjoyable, you're not going to do it either, are you? Well, I've got news for you: neither will your dog. So be sure to make training fun.
Lure/reward training is expedient
Pet dog training techniques must be appropriate and suitable for all pet dog owners (men, women, and children) to train all types of pet dog-all breeds, all mixes, and all shapes and sizes.
Like most experts, dog trainers severely underestimate their own expertise and experience. Indeed, most dog training books describe complicated and time-consuming techniques, which the author (usually an adept and experienced trainer) uses to train specific breeds of dog (such as Golden Retrievers, Labrador Retrievers, Border Collies, German Shepherds, and Belgian Malinois), that have been selectively bred for their ease and excellence for competitive obedience, working trials, or protection work. Many of these methods and techniques are largely unsuitable for inexperienced trainers (most pet owners) and for many breeds and mixes of dog (especially quick little dogs, big cumbersome dogs, fearful dogs, and dogs with attitude).
The owner-friendly and dog-friendly dog training techniques outlined in this little book are gentle, easy, efficient, effective, efficacious, expedient, and extremely enjoyable for you and you dog. And so without further ado, turn to Chapter 4, "Good Manners," and begin training your puppydog right away. Lure/reward training techniques work like a treat!
Excerpted from Doctor Dunbar's Good Little Dog Book, by Ian Dunbar.
Ian Dunbar is a veterinarian and animal behaviorist, founder of the Association of Pet Dog Trainers, and the author and star of numerous books and videos on dog behavior and training. He lives in Berkeley, California with his wife, trainer Kelly Dunbar, and their three dogs. The Dunbars are contributing editors to DogTime.
latest news & articles
offers from our sponsors