By Jeff Stallings, CPDT-KA
The Bay Area’s Michael Pollan is one of my favorite non-fiction authors. His books about the history and culture of food production and consumption are compelling and wide-ranging,
One thought from Pollen’s 2007 book “In Defense of Food” gained notoriety for its simplicity and truth:
- Eat food.
- Not too much.
- Mostly plants.
So simple and yet so powerful!
I’d like to borrow the cadence from this simple set of phrases to illustrate that using treats to teach your dog what’s expected is by far the most effective tool in your dog training toolbox:
- Feed treats.
- Not too many.
- Mostly trade.
Vast swaths of dog trainers and their clients understand that positive reinforcement is the best way to teach dogs desirable behaviors, but I still encounter people who refuse and say, “I don’t want to bribe my dog”. Years ago, when I first got into dog training as a profession, the trainer I briefly apprenticed with made the argument that “wolves don’t dig up bones to bribe their pups to hunt.” Such a lame analogy!
Dogs are not wolves!
First off, dogs haven’t been wolves for 30,000 years and their behavior differences are immense. Also, while dogs (or wolves) can learn from each other, this is called “social learning” and is not the same as positive reinforcement, which requires well-timed rewards for behaviors you would like repeated in the future.
Second, the term “bribe” implies cheating by unscrupulous means to get someone to do something unsavory. Implementing positive reinforcement using food to show your dog behaviors you want them to perform on cue is anything BUT cheating: It’s communication. And not only communication, but interspecies communication. What could be more amazing than that?
When I hear someone say “I don’t want to bribe my dog”, my strategy is to ask them if they would continue to show up for work if the paychecks stopped. Invariably the reply is “no”, so then I ask why their dog should be expected to work for free. It’s all semantics, but if helps for you think of the food rewards during positive reinforcement training as a paycheck rather than a bribe, then a paycheck it is!
So, let’s break it down:
Feed treats
Humans have many things that could be considered “primary reinforcers”, things that are intrinsically valuable. Food can be a reinforcer, as can sex, paychecks, new iPhones, all-expenses paid vacations. I could probably get you do to many things you otherwise would not in exchange for a new car. For humans, the list of primary reinforcers is long.
Primary reinforcers for dogs (and all animals) are biologically important: food, sleep, shelter, safety and sex. The relevant primary reinforcer for this discussion is food, one we have in common with our dogs: I don’t have to teach you to desire and seek out chocolate, nor do we have to teach our dogs to seek out (or work for) treats. Forgo food for too long, you die. That’s essentially what makes a reinforcer primary.
Not too many
Whenever I start working with a new client, one concept we nail down right off the bat is effective use of reinforcement schedules. What are reinforcement schedules, you may ask? I wrote a blog post that covers this in more detail, but here’s the elevator talk:
When you begin teaching a new command, for instance training a young puppy to sit, you give the puppy a treat each time he successfully performs the behavior. You are like a vending machine: Puppy sits, treat is given. Ergo puppy learns that sitting when they hear the word “sit” causes good things (food) to occur.
The problem happens when this human-as-vending-machine goes on for too long. If you continue to give your puppy a treat every time he sits, he learns that he is to perform the command when (1) he hears the word “sit” and (2) a treat is present. But you want the behavior to become (literally) second nature, for your dog to perform the behavior whether you happen to have a treat or not. This is where the “not too many” piece comes in.
As soon as your pup reliably performs the behavior on cue, it’s time for you to shapeshift from vending machine to slot machine, to NOT give a treat every time, to keep your puppy guessing. At first you may reward with treats 50% of the time, then roughly 30% of the time. Sometimes he gets three in a row, sometimes none for six iterations; you are a slot machine, after all. And slot machines are far more interesting than vending machines!
Mostly trade
Another handout I give my training clients is entitled “Nothing in Life is Free”. The gist is that every interaction with your dog is a training opportunity, not just those moments you have a treat pouch on your belt and a clicker in hand. This is about building communication between you and your dog by having him “work” in exchange for anything he finds valuable including play, toys, walks, belly rubs, access to the couch and, of course, food. When you give your dog their dinner without asking for even a simple “sit” in exchange, you are wasting a training opportunity.
My dog Otis is 11 years old and incredibly well trained. (You would hope, no?) But even at this age, when I give her a frozen kong to eat while I watch a movie, I have her perform some of the myriad commands we have built together. She sits (obviously), lays down, weaves between my legs, gives me a fist bump (my fist/her nose), and so on. Then and only then does she get her Kong. Is this a bribe? Heck no, it’s a paycheck and one she loves to earn.
So that sums it up: Feed treats. Not too many. Mostly trade.
Treats or toys?
One caveat here: Some dogs prefer to “work” for a toy and play than receive food. You have to find out what motivates your dog in the moment. Otis is half border collie, and that half would engage me in play 24/7 if I let her. Food is the most reliable reinforcer for the vast majority of dogs, but some, especially herding breeds, would rather work for play than food. And that is absolutely fine. In fact, I would argue that for working breeds, the work IS a primary reinforcer, one related to the hunting instincts in their wolf forbears.
So give your dog treats as a paycheck (not a bribe), don’t overdo it, and have him perform commands in exchange for what he finds valuable. You’ll both be set for years of fun, exciting interspecies communication!
