SIT, SIT, SIT!

Have you ever paused to consider what criteria you have for a “well-trained” dog? I have. As a dog trainer, I have had thousands of appointments with clients who tell me about how well-trained their dogs are, only to turn around and nag, beg, or cajole their dogs into obedience. I have had people insist that their dogs come when called 80% of the time or more, only to have the dog refuse to come when called even in the most distraction free zone we can find. Sit, sit, sit, no, Fluffy, SIT! is the litany for these people, only to have Fluffy sit and stare up at them in disbelief as though they are asking for the moon. Sitting is not difficult for most dogs to perform, but it sure seems difficult for many dogs to perform when asked in my presence.

This is a great picture of a well trained dog. The dog and the man are leaning into one another, and they are obviously engaged with one another. I doubt that the dog just randomly sat on the middle of a bridge for a stranger, so that tells me he probably sat because he was asked, not because he was forced to or promised a cookie. He and the man are simply working together.

So what do I mean when I say “I have a well-trained dog”? I mean that I have a deep connection with the dog and we work together. Often the dog will perform for long duration in the absence of either treats or a threat. This means that in general, if I am out and about with my well-trained dog, and I need him to sit, he will sit on the first calm quiet and gentle request, and I don’t have to wave a treat in front of him, nor do I have to physically force him into position nor yank on a collar. I don’t have to loom, lean in or raise my voice. Furthermore, in the event that it is not a good idea from my dog’s perspective and he refuses, his refusal is understood to be reasonable from his point of view. Let’s parse all my conditions one at a time.
When I say that my dog will perform a behaviour on the first ask, and that I need only ask in a calm, quiet and gentle way I am thinking here about what it might look like when I ask someone to pass the salt. If we are sitting at dinner and I would like some salt, I might look at you, make eye contact, say your name and then calmly say “please pass the salt”. Pass the salt is generally a reasonable request at the dinner table, and it doesn’t take a lot of effort for you to respond and my expectation is that you will hear, attend to my request and just pass me the salt. In other words, not a big deal. So if I ask my dog to sit, and he is well-trained, I can just ask, and he will just sit.

If we are eating at the same table and I ask you to pass the salt, I don’t usually need to raise my voice, threaten you, promise you a reward or loom over you. I can simply ask you to pass the salt. Well trained dogs are like this. It should be easy.

Let’s consider my “no treat” criteria. It is well understood that behaviours need to be reinforced in order to be built and maintained. The thing is that once a behaviour is very strong with a reinforcer of some sort, you need to randomize the reinforcement in order to maintain the behaviour. It is a bit like a slot machine compared to a pop machine. When you put a coin into a pop machine, you expect that a pop will be produced. You expect that this will happen every single time. This is different than a slot machine, where you are expected to put the coin in and pull the lever in order to get nothing! You might someday get something, if you are exceptionally lucky, but the chance of getting something big is really low. Sit as a behaviour works a little like this. Initially, when I am teaching a dog to sit, I am a pop machine. Every single time that the dog sits, I give him something he wants. At first this will be something like the pop machine. I am careful about two things at that stage. First is that I don’t show the dog what the reinforcer will be. Sometimes I use steak, or a special dog training sausage, or some cheese, or the chance to chase a ball or go through a door or get in a car, but I am not going to hang a piece of bacon off my hat in the hopes that the dog will perform based on a bribe. Bribes are technically reinforcers that are presented ahead of the behaviour and eventually, they will backfire on you. The dog will look for the reinforcer and measure if the offered item is worth what you are asking for.

Showing the dog a treat BEFORE you ask for the behaviour is called bribery. A well trained dog should not need to be bribed to perform a behaviour.

Once I am sure that the dog knows that the behaviour of sit will be rewarded, I pull a little switch on the dog. I ask for the sit and I “goof”. Before I do this, I have to be really, really sure that my dog both understands what I mean when I quietly and calmly and gently ask for the behaviour. I need to have practiced this behaviour with a guarantee of a reward in a wide variety of places. Once I am really sure, I pretend that I forgot the reward. I might move away, and then ask the dog to perform the behaviour again. If the dog is convinced that I am reliable and that I just goofed, he will offer the behaviour a second time and usually, he will offer it with more intensity and vigour. The second time I ask, I throw a party and give the dog something EXTRA special. I might even goof again and give him his reward twice. The key here is to make that second sit, more rewarding than he thought the first one would be. Then I go back to rewarding every sit for a bit and then I “goof” again. I keep doing this until the times I goof are no big deal. Then I start goofing a little more often. I keep doing this until I have built a truly random level of goofs, but where the dog is still keen to do the behaviour. I am still randomizing what the reward is for the dog though. It is important that the dog is keen on the game. Eventually I have thinned things out so much that the dog is sitting for free more often than he is being paid but when he is paid he gets something really special.
A well-trained dog does not need a threat in order to work. One of the saddest trends I see from time to time in the training world is the idea that you have to deepen your voice and increase the volume in order to get the dog to do as he is asked. If you need to do this in order to get your dog to behave as you wish, you are not on the same team as your dog. Does this mean I have never raised my voice to my dog? Of course not; if the stakes are high, and his life is at risk, then yes, I would raise my voice and insist, but if I think I have to do this all the time, what does that say about my relationship with my dog? Not anything terribly pleasant! If you find yourself doing this, you should probably go back to reinforcing absolutely every sit.
Along with a threatening voice comes this issue of physically pushing the dog into a behaviour, or yanking on the leash to make the behaviour happen. When I am training a dog there are a few limited instances of when I might choose to use a physical penalty with a dog but in each case that is to stop a behaviour altogether, not when I need to make him start to do something. If a dog were eating something that could cause him harm such as a rock, I might grab him by the collar for instance, and try to startle the dog out of the behaviour, but I don’t use it to try and initiate a behaviour. Most of the time, when a dog is asked to do a behaviour, such as sit, and he refuses, if you jerk on his neck with the leash, you are just going to teach him to guess what you want. He isn’t going to know what you want. The most common effect of that sort of training is a dog who is slow because they are concerned that they are going to get hurt, even just a little bit, and they are afraid to act.

This dog is obviously afraid of something and the most likely thing is a physical prompt of some sort. Why do I say that he looks afraid? His ears are pinned and his face is tense and his feet are clenched. His back is roached too! This is not a well trained dog.

Another type of threat I see is looming over or leaning over the dog, and repeating the request in a louder or more demanding voice. Again, if you have to do this, you really need to go back to the beginning and start over. A well-trained dog doesn’t need you to loom over him or raise your voice. When this happens I often think of a scene from The Addams Family movie where the little girl, Wednesday, asks her uncle to pass that salt. Her mother, Morticia, sweetly prompts her, saying “Wednesday, what do we say?” and of course most of us think “please?” but Wednesday turns to her uncle, raises her voice and demands “NOW!” If you need to demand compliance from your dog, he doesn’t know the behaviour sufficiently well in order to be successful often enough to actually be well-trained.
To me the final stage of having a well-trained dog is to have a dog who has the possibility of refusing a behaviour when it is not a good idea for him to do that. This implies that if I ask for something reasonable, such as sit, if the dog refuses there is likely a pretty good reason for that. When I have a young puppy who seems to know sit refuse to sit, I find that the pup often needs to toilet. Sitting with a full bladder or bowel is uncomfortable. Forcing the puppy to sit under these circumstances is unkind and can interrupt house training. If I have a middle-aged dog who is very reliable on sit, and I ask for it and he doesn’t comply, I watch him move, because often the root cause is pain. Sometimes though, when an adult dog won’t sit, he may have another reason not to do as I ask. He may be afraid and feel vulnerable, or he may not want to sit down on something that is uncomfortable. I had one dog refuse to do a down stay, his strongest behaviour of all, because I left him on a bee’s nest. He stood up, moved three feet, and lay back down. I was really upset until I realized where I had left him. Sadly, this was during an obedience competition and he lost our class! I cannot get upset at a dog who had the sense to move himself before he got stung by bees. Luckily I had enough trust in him to trust that he would not behave randomly, and he had enough trust in me to change his position and keep himself from harm. I know of dogs who would have taken the stings of a nest of bees instead of moving because they had been punished severely enough for moving that bees were not a bad enough pain! I have also met trainers who were really proud of the fact that their dogs had been willing to take that pain instead of moving. That does not speak well to the depth of the relationship that the person has with the dog, and that is unfortunate. At the end of the day, I value my relationship with my dog more than I value instant and perfect obedience.

A well trained dog should be willing to perform the trained behaviour anywhere, anytime, under many different circumstances. This dog is wet, and the bench is likely not very comfortable to sit on, but he is willing to sit and stay for the picture.

As a final point, not included above, I feel that a well-trained dog should know more than just one behaviour and he should be able to perform those behaviours in a wide variety of situations and circumstances. Being able to perform them in only one place or under only one circumstance does not reflect good training. It should not matter to the dog if you ask him for a behaviour while you are standing in front of him, or beside him, or on the landing of your deck, or if you are sitting in a chair. He should understand the behaviours in a wide variety of situations, and of course that means practicing, from the beginning, everything your dog is being trained to do.
So when I think about a well-trained dog, I am not thinking about a little canine robot, but I am also not thinking of a dog who is completely on his own agenda either. I am looking for evidence that the dog and handler are working together, in partnership, as a team. I am looking for evidence that the human is helping the dog without luring or bribing the dog. I am looking for evidence that the dog is thinking things through and not performing the behaviour solely to get the reward or to avoid some harsh penalty. That is what I mean by a well-trained dog.