Why Do We Train With Food?

This article by Dr. Cynthia D. Fisher, B.A. MS, PhD - Chief Instructor (Obedience) at Gold Coast Dog Obedience Club - is reproduced with the kind permission of the author.


Occasionally we encounter a handler who is uneasy with the idea of using food in training their dog. For many years, the competition obedience community frowned on the use of food. It was seen as somehow cheating. Instead, people trained with praise and petting as their rewards, and used a lot of physical corrections (like checks on a chain collar or pushing the dog into the sit or drop position) to show the dog what they wanted or to punish mistakes.

Today, many trainers have realised that punishment based training is out-of-date, unfair to the dog, and less effective than reward-based training. The past decade has seen a revolution in positive dog training techniques around the world, and our club now uses these state of the art methods.

Psychologists have known the basic principles of animal learning that we apply in positive dog training since the 1940s. In the 1970s, marine mammal trainers become very good at using the principles of positive training. They had to, because they couldn't put a collar on a killer whale or a dolphin and FORCE it to do anything!

It's been said that we have historically used coercive training methods on dogs because we could. Dogs are land-based animals that are smaller than we are and generally prefer not to attack us, so we have been able to get away with harsh methods. But these methods are not necessary, and are not particularly effective. It is now possible to train your dog without even touching it. If you are smarter than your dog, you no longer need to be tougher and meaner than it is to get it to do what you want!

All animals (and people) learn to repeat behaviours that have had pleasant consequences. A pleasant consequence is called a positive reinforcer. Training is all about getting the animal to do what you want and then providing an immediate positive reinforcer so the animal learns to do that behaviour again in the future. There are many possible reinforcers for dogs: praise, petting, play (with a ball or tug toy or other game the dog likes), food treats, and life rewards.

Life rewards are any desirable privilege you are likely to give your dog anyhow in the course of a day, like letting it in the house, taking it for a drive or a walk, inviting it up on to the sofa, etc. If the dog likes the reward, it will work to attain it.

Remember, REINFORCERS ARE IN THE EYE OF THE BEHOLDER. It's the dogs opinion that counts. There is no point in trying to train a dog using rewards that the dog doesn't want or find reinforcing. If your dog isn't interested in kibble (dried dog food), try cheese or meat. If your dog doesn't like being patted on the head (and most don't), don't try to use pats on the head as a reinforcer because the dog won't work to earn more of them.

FOOD IS NOT QUALITATIVELY DIFFERENT than petting or praise - it's just another reinforcer, another kind of positive consequence. But food happens to be a reinforcer that has much greater value than praise or petting for most dogs. Dogs need food to live, so nature equipped them to value it highly. They don't need praise or petting to live.

So why not train with a potent reinforcer like food rather than a weaker reinforcer like petting? After all, you wouldn't dig a swimming pool in your back yard with a teaspoon; you'd get a steam shovel and get the job done quickly and effectively!

Training with food is faster and more powerful than training with weaker rewards like praise and petting. And you can, of course, always use praise and touch along with food.

There are a number of other good reasons to use food in training:

Using food motivates the dog to pay attention to you, and attention is a prerequisite to almost any learning or training. When you are training your dog, you are competing against the environment for its attention. On the training grounds, there are many other dogs, lots of people, and plenty of interesting scents.

You need to be more interesting to your dog than all this. Your chances are better if you've got sausage than if you've got praise. One problem with praise and petting as reinforcers is that WE GIVE THEM AWAY FOR FREE much of the time. We stroke our dogs and tell them how wonderful they are because we like them, day in and day out. This is one of the joys of owning a dog. But why should they work for something they get plenty of for free?

Some people who have success training dogs with praise and petting intentionally deprive their animals of human contact by kennelling them at all times other than while training. This may increase the animals desire for praise and petting, but it's not much of a life for a pet.

It is completely natural for dogs to work for food. The wild dog or wolf certainly has to work pretty hard to track down and kill something to eat! Zoos often invent food puzzles so their captive animals have to work and use their brains and muscles to obtain food. This mental and physical stimulation is very beneficial to their health and well-being. Food training gives your dog the same enriching experience.

Being the source of food increases your status and leadership in the dogs eyes, especially as you set the rules for getting the food. Why should the dog look to you for leadership if it appears that God fills a bottomless food bowl each day, and you've got relatively little of real value to offer the dog contingent on its behaviour?

Training with rewards makes the dog eager to do more, to work harder, and to pay attention to you. Training with punishments makes the dog want to do just enough to escape or avoid punishment, and may greatly reduce its desire to be with you.

Dogs learn by reinforced repetition. It is easy to practice an exercise ten times in a row if you reward each repetition with a tiny piece of food. It's harder to get a training rhythm going and do ten repetitions if your reward is throwing a ball. It takes a while to get the dog and ball back and set up for another repetition. It's even harder to do many repetitions in a short period if your reward is taking the dog for a walk in the park. This is not to say you shouldn't use ball games or walks in the park as rewards, by all means do so! But when teaching a new behaviour, it's repetitions that produce learning, and most dogs will do more repetitions for food in a short while than for other reinforcers.

Eating has a calming effect on dogs. It does on us too- we sometimes eat to relieve stress in our lives. Dogs that are anxious or aggressive tend to calm down when food is used in training. These same dogs become more anxious or aggressive when harsh methods add extra stress to training.

Food can be used in the early stages of training as a lure as well as a reinforcer. If you lure your dog into the sit or drop position with food, you don't need to physically push or pull or force it into that position. And force from us tends to cause resistance from dogs. With a food lure, the dog freely chooses to take up the sit or drop position.

If you regard your dog as your best friend, you probably don't want to jerk it around roughly or cause it pain if the was another way to teach it to listen to you and behave well. AND THERE IS!